Thermodynamics, kinetics, and mechanism of (silox)3M(olefin) to (silox)3M(alkylidene) rearrangements (silox = tBu 3SiO; M = Nb, Ta)

Published on Web 03/10/2005

Thermodynamics, Kinetics, and Mechanism of

(silox)₃M(olefin) to (silox)₃M(alkylidene) Rearrangements

t

(silox ) Bu₃SiO; M ) Nb, Ta)

Kurt F. Hirsekorn,^†Adam S. Veige,^†Michael P. Marshak,^†Yelena Koldobskaya,^†

Peter T. Wolczanski,*^,†Thomas R. Cundari,^‡and Emil B. Lobkovsky^†

Contribution from the Department of Chemistry and Chemical Biology, Baker Laboratory,

Cornell UniVersity, Ithaca, New York 14853, and Department of Chemistry, UniVersity of North

Texas, Box 305070, Denton, Texas 76203-5070

Received June 28, 2004; E-mail: ptw2@cornell.edu

Abstract: Olefin complexes (silox)₃M(ole) (silox ) ^tBu₃SiO; M ) Nb (1-ole), Ta (2-ole); ole ) C₂H₄, C₂H₃-

t

Me, C₂H₃Et, C₂H₃C₆H₄-p-X (X ) OMe, H, CF₃), C₂H₃Bu, ^cC₅H₈, ^cC₆H₁₀, ^cC₇H₁₀(norbornene)) rearrange to

alkylidene isomers (silox)₃M(alk) (M ) Nb (1dalk), Ta (2dalk); alk ) CHMe, CHEt, CHⁿPr, CHCH₂C₆H₄-

p-X (X ) OMe, H, CF₃(Ta only)), CHCH₂Bu, ^cC₅H₈, ^cC₆H₁₀, ^cC₇H₁₀(norbornylidene)). Kinetics and labeling

t

experiments suggest that the rearrangement proceeds via a δ-abstraction on a silox CH bond by the â-olefin

n

carbon to give (silox)₂RM(κ²-O,C-OSi^tBu₂CMe₂CH₂) (M ) Nb (4-R), Ta (6-R); R ) Me, Et, Pr, Bu, CH₂-

t

c

CH₂C₆H₄-p-X (X ) OMe, H, CF₃(Ta only)), CH₂CH₂Bu, C₅H₉, C₆H₁₁, C₇H₁₁(norbornyl)). A subsequent

R-abstraction by the cylometalated “arm” of the intermediate on an R-CH bond of R generates the alkylidene

1dalk or 2dalk. Equilibrations of 1-ole with ole′ to give 1-ole′ and ole, and relevant calculations on 1-ole

and 2-ole, permit interpretation of all relative ground and transition state energies for the complexes of

either metal.

ous cross-metathesis methodologies¹⁵are forefront in catalytic

applications to the preparation of fine chemicals.

Introduction

While the field of organometallic chemistry boasts a number

of unique processes, the olefin metathesis reactionsand its

heterogeneous analoguesis perhaps the most widespread in

terms of application.¹Commodity chemicals, polymers, and fine

chemicals are all prepared via judicious use of alkene

metathesis.^1-8Surfactant and plasticizer production via the shell

higher olefins process employs olefin cross metathesis to

generate aldehyde and alcohol precursors from internal olefins

that are too short or long.^2-5,9,10The disproportionation of

propene in the Philips triolefin process affords butene and

ethylene, which is the more useful alkene.^2-5Extension of the

olefin metathesis reaction to polymer synthesis has established

ring-opening metathesis polymerization and acyclic diene

metathesis^2,6as attractive, relatively new approaches to highly

functionalized olefin-containing polymers. Finally, ring-closing

metathesis,^11-14including enantioselective variants,¹²and vari-

Once the Chauvin mechanism¹⁶for olefin metathesis was

established,^17-21the key discovery in the development of the

process was the synthesis of stable metal alkylidene complexes,

t

i.e., L_nMdCRR′ (R, R′ ) H, aryl, Bu, etc.), that served as

catalysts or catalyst precursors. Schrock’s seminal synthetic

work, and the advent of R-abstraction as a synthetic tool,^21-23

enabled early transition metal chemistry to showcase olefin

(11) (a) Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413-4450. (b) Grubbs,

R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995, 28, 446-552.

(12) (a) Schrock, R. R.; Hoveyda, A. H. Angew. Chem., Int. Ed. 2003, 42, 4592-

4633. (b) Tsang, W. C. P.; Jernelius, J. A.; Cortez, G. A.; Weatherhead,

G. S.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2003, 125; 2591-

2596. (c) Schrock, R. R.; Jamieson, J. Y.; Dolman, S. J.; Miller, S. A.;

Bonitatebus, P. J., Jr.; Hoveyda, A. H. Organometallics 2002, 21, 409-

417.

(13) Schuster, M.; Blechert, S. Angew. Chem., Int. Ed. Engl. 1997, 36, 2036-

2056.

(14) Roy, R.; Das, S. K. Chem. Commun. 2000, 519-529.

(15) (a) Chatterjee, A. K.; Choi, T. L.; Sanders, D. P.; Grubbs, R. H. J. Am.

Chem. Soc. 2003, 125, 11360-11370. (b) Blackwell, H. E.; O’Leary, D.

J.; Chatterjee, A. K.; Washenfelder, R. A.; Bussmann, D. A.; Grubbs, R.

H. J. Am. Chem. Soc. 2000, 122, 58-71.

^†Cornell University.

^‡University of North Texas.

(1) (a) Rouhi, A. M. Chem. Eng. News 2002, 80, 29-33. (b) Rouhi, A. M.

(16) Herrison, J. L.; Chauvin, Y. Makromol. Chem. 1970, 141, 161.

(17) Katz, T. J. AdV. Organomet. Chem. 1978, 16, 283-317.

(18) Calderon, N.; Lawrence, J. P.; Ofstead, E. A. AdV. Organomet. Chem. 1979,

17, 449-492.

Chem. Eng. News 2002, 80, 34-38.

(2) Grubbs, R. H., Ed. Handbook of Metathesis; Wiley & Sons: New York,

2003.

(3) Ivin, K. J.; Mol, J. C. Olefin Metathesis and Metathesis Polymerization;

Academic Press: New York, 1997.

(19) (a) Love, J. A.; Sanford, M. S.; Day, M. W.; Grubbs, R. H. J. Am. Chem.

Soc. 2003, 125, 10103-10109. (b) Dias, E. L.; Nguyen, S. T.; Grubbs, R.

H. J. Am. Chem. Soc. 1997, 119, 3887-3897.

(4) Parshall, G. W.; Ittel, S. D. Homogeneous Catalysis; Wiley & Sons: New

York, 1992.

(20) Adlhart, C.; Hinderling, C.; Baumann, H.; Chen, P. J. Am. Chem. Soc.

2000, 122, 8204-8214.

(5) Mol, J. C. J. Mol. Catal. A; Chem. 2004, 213, 39-45.

(6) Buchmeiser, M. R. Chem. ReV. 2000, 100, 1565-1604.

(7) Fu¨rstner, A. Angew. Chem., Int. Ed. 2000, 39, 3012-3043.

(8) Mol, J. C. Green Chem. 2002, 4, 5-13.

(9) Freitas, E. R.; Gum, C. R. Chem. Eng. Prog. 1979, 75, 73-76.

(10) Reuben, B.; Wittcoff, H. J. Chem. Educ. 1988, 65, 605-607.

(21) Schrock, R. R. J. Mol. Catal. A: Chem. 2004, 213, 21-30.

(22) (a) Schrock, R. R. Acc. Chem. Res. 1979, 12, 98-104. (b) Schrock, R. R.

Pure Appl. Chem. 1994, 66, 1447-1454. (c) Feldman, J.; Schrock, R. R.

Prog. Inorg. Chem. 1991, 39, 1-74. (d) Schrock, R. R. Acc. Chem. Res.

1990, 23, 158-165.

9

J. AM. CHEM. SOC. 2005, 127, 4809-4830

4809

A R T I C L E S

Hirsekorn et al.

metathesis in detailed mechanistic studies and catalytic applica-

tions. More recently, the dramatic increase in functionality

tolerance exhibited by Grubbs’ catalysts^{2,11,15,19,24}and variants⁷

and Schrock’s creative exploitation of molybdenum^12,21have

exponentially increased the use of olefin metathesis in fine

chemicals synthesis.^7,11-15

Early in the history of alkylidene development, it was

recognized that rearrangement of L_nMdCR(CH₂R′′) to an olefin

complex L_nM(RHCdCHR′′) could be a potentially damaging

process in relation to metathesis catalysis. In fact, the seeming

inability to synthesize alkylidenes with â-CH bonds was often

blamed on their intrinsic instability with respect to a bound

olefin, i.e., K_alk/ole< 1 in eq 1.²⁵While most studies regarding

number of olefin complexes have been prepared, but until

recently, none had been subjected to high temperature ther-

molysis. Preliminary indications with (silox)₃Nb(ole) (1-ole, ole

) 1-butene, cyclohexene) and [(silox)₃Nb]₂(η-1,2;η-5,6-^cC₈H₆)

indicated that K_alk/ole> 1 at elevated temperatures.³⁶Moreover,

little olefin metathesis activity was noted, despite ligands that

would be typically expected to support such reactivity. Given

these tantalizing examples, and the prospect that the sterics

intrinsic to the (silox)₃M (M ) Nb, 1; Ta, 2) core would permit

interrogation of the olefin to alkylidene rearrangement without

interference from olefin metathesis, a systematic study of eq 1

was conducted, and it is reported herein.

Results

Synthesis of (silox)₃Nb(olefin) Complexes. Two methods

were used to prepare niobium olefin complexes. The most

practical preparation of (silox)₃Nb(ole) (1-ole, ole ) olefin)

involved Na/Hg reduction of (silox)₃NbCl₂(3) in the presence

of an excess of olefin, typically with THF as the solvent (eq

2).^36,37

olefin metathesis appear to support this premise, a limited

number of specific mechanistic studies have been attempted.

An investigation of cationic rhenium complexes by Gladysz and

Hatton led to an interpretation of the system (K_alk/ole< 1) as an

organometallic Wagner-Meerwein rearrangement in reference

to its carbocation-like hydrogen migration.²⁶A study by Bercaw

et al. of cyclometalated tantalum olefin and alkene complexes

showed only a modest thermodynamic preference for the latter.²⁷

A significant number of observations suggest that a blanket

statement pertaining to the instability of alkylidenes with

â-hydrogens is dogmatic. Schrock et al. have catalyzed the

rearrangement of ethylene to ethylidene via addition of PhPH₂,²⁸

and there are several other systems in which a greater

thermodynamic stability of the alkylidene is implicated by

certain reactivity sequences,^29-35including clear examples of

olefin to alkylidene and alkylidyne rearrangements by Caulton

et al.³⁴

THF, 24 h

_{Na/Hg, -2 NaCl}8

(silox)₃NbCl₂+ olefin (excess)

3

(silox)₃Nb(ole)

(2)

(3)

1-ole (ole ) ^cC₇H₁₀, 25%; ^cC₆H₁₀, 40%)

+ olefin (excess) ^benzene8

(silox) Nb(4-pic)

3

1-4-pic

(silox)₃Nb(ole) + 4-picoline

1-ole (ole ) C₂H₄, C₂H₃Me (67%); C₂H₃Et (81%);

cis-MeCHCHMe (40%); H₂CCHC₆H₄-p-X,

X ) H (44%), OMe (46%), CF₃(46%)

Yields are modest (∼40%), but the synthesis is direct as long

as an excess of olefin can be used. An alternative methodology

While the portrayal of â-hydrogen-substituted alkylidenes as

intrinsically unstable is compelling in view of the limited number

of stable examples, L_nMdCR(CH₂R′′) species must be inter-

mediates in a variety of catalytic applications,^1-24and their

potential to rearrange does not appear to be a major stumbling

block to utilization. Several questions remain. First, are L_nMd

CR(CH₂R′′) species thermodynamically unstable with respect

to their olefin congeners or are K_alk/oledependent on metal or

substrate? Second, are these rearrangements kinetically swift,

or are there substantial impediments to the rearrangement

process?

(29) Miller, G. A.; Cooper, N. J. J. Am. Chem. Soc. 1985, 107, 709-711.

(30) Hughes, R. P.; Maddock, S. M.; Rheingold, A. L.; Guzei, I. A. Polyhedron

1998, 17, 1037-1043.

(31) Giannini, L.; Guillemot, G.; Solari, E.; Floriani, C.; Re, N.; Chiesi-Villa,

A.; Rizzoli, C. J. Am. Chem. Soc. 1999, 121, 2797-2807.

(32) (a) Schrock, R. R.; Seidel, S. W.; Mosch-Zanetti, N. C.; Shih, K.-Y.;

O’Donoghue, M. B.; Davis, W. M.; Reiff, W. M. J. Am. Chem. Soc. 1997,

119, 11876-11893. (b) Schrock, R. R.; Seidel, S. W.; Mosch-Zanetti, N.

C.; Dobbs, D. A.; Shih, K.-Y.; Davis, W. M. Organometallics 1997, 16,

5195-5208.

(33) (a) Kleckley, T. S.; Bennett, J. L.; Wolczanski, P. T.; Lobkovsky, E. B. J.

Am. Chem. Soc. 1997, 119, 247-248. (b) Kleckley, T. S., Ph.D. Thesis,

Cornell University, Ithaca, NY, 1998.

(34) (a) Ozerov, O. V.; Pink, M.; Watson, L. A.; Caulton, K. G. J. Am. Chem.

Soc. 2004, 126, 6363-6378. (b) Ozerov, O. V.; Pink, M.; Watson, L. A.;

Caulton, K. G. J. Am. Chem. Soc. 2004, 126, 2105-2113. (c) Ozerov, O.

V.; Pink, M.; Watson, L. A.; Caulton, K. G. J. Am. Chem. Soc. 2003, 125,

9604-9605.

In examining the chemistry of (silox)₃NbL (1-NbL, L ) pyr,

4-pic,³³PMe₃)^36-38and (silox)₃Ta (2)^39-44over the years, a

(35) (a) Carmona, E.; Paneque, M.; Poveda, M. L. J. Chem Soc., Dalton Trans.

2003, 4022-4029. (b) Padilla-Martinez, I. I.; Poveda, M. L.; Carmona,

E.; Monge, M. A.; Ruiz-Valero, C. Organometallics 2002, 21, 93-104.

(36) Veige, A. S.; Wolczanski, P. T.; Lobkovsky, E. B. Angew. Chem., Int. Ed.

2001, 40, 3629-3632.

(23) (a) Schrock, R. R. Chem. ReV. 2002, 102, 145-179. (b) Schrock, R. R. J.

Chem. Soc., Dalton Trans. 2001, 2541-2550. (c) Wallace, K. C.; Liu, A.

H.; Dewan, J. C.; Schrock, R. R. J. Am. Chem. Soc. 1988, 110, 4964-

4977.

(24) Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001, 34, 18-29.

(25) (a) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles

and Applications of Organotransition Metal Chemistry; University Science

Books: Mill Valley, CA. (b) Mingos, D. M. P. In ComprehensiVe

Organometallic Chemistry; Wilkinson, G., Stone, F. G. A., Abel, E. W.,

Eds.; Pergamon: New York, 1982; Vol. 3, Chapter 19.

(37) Veige, A. S.; Kleckley, T. S.; Chamberlin, R. L. M.; Neithamer, D. R.;

Lee, C. E.; Wolczanski, P. T.; Lobkovsky, E. B.; Glassey, W. V. J.

Organomet. Chem. 1999, 591, 194-203.

(38) (a) Veige, A. S.; Slaughter, L. M.; Lobkovsky, E. B.; Wolczanski, P. T.;

Matsunaga, N.; Decker, S. A.; Cundari, T. R. Inorg. Chem. 2003, 42, 6204-

6224. (b) Veige, A. S.; Slaughter, L. M.; Wolczanski, P. T.; Matsunaga,

N.; Decker, S. A.; Cundari, T. R. J. Am. Chem. Soc. 2001, 123, 6419-

6420.

(39) Chadeayne, A. R.; Wolczanski, P. T.; Lobkovsky, E. B. Inorg. Chem. 2004,

43, 3421-3432.

(40) Neithamer, D. R.; LaPointe, R. E.; Wheeler, R. A.; Richeson, D. S.; Van

Duyne, G. D.; Wolczanski, P. T. J. Am. Chem. Soc. 1989, 111, 9056-

9072.

(26) (a) Roger, C.; Bodner, G. S.; Hatton, W. G.; Gladysz, J. A. Organometallics

1991, 10, 3266-3274. (b) Hatton, W. G.; Gladysz, J. A. J. Am. Chem.

Soc. 1983, 105, 6157-6158.

(27) Parkin, G.; Bunel, E.; Burger, B. J.; Trimmer, M. S.; Van Asselt, A.;

Bercaw, J. E. J. Mol. Catal. 1987, 41, 21-39.

(28) Freundlich, J. S.; Schrock, R. R.; Davis, W. M. J. Am. Chem. Soc. 1996,

118, 3643-3655.

9

4810 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

A R T I C L E S

involved treatment of (silox)₃Nb(η-N,C-4-pic) (1-4-pic, 4-pic

) 4-picoline) with a slight excess of olefin (g10 equiv in the

cases of cyclohexene and norbornene).³⁶A directly related

method was used previously in the synthesis of (silox)₃Nb(η-

H₂CCHPh) (1-C₂H₃Ph) from (silox)₃Nb(η-N,C-NC₅H₅).^34,37In

cases where the olefin is bulky or otherwise coordinates poorly,

the 4-picoline can be a competitive binder. For example, no

indication of 4-picoline displacement was evidenced when 1-4-

pic was exposed to >1 equiv of trans-2-butene, but the addition

of 5.0 equiv of cis-2-butene afforded (silox)₃Nb(η-cis-C₂H₂-

Me₂) (1-c-2-butene), although some 1-4-pic remained (∼7%,

K(eq 3) ≈ 3.0). Alternatively, (silox)₃NbPMe₃(1-PMe₃)³⁸could

be used as a source of “(silox)₃Nb” (1), but the expense of this

reagent limited its application to a few olefin adducts, as eq 4

indicates.

₃+ olefin (excess) ^pentane8

(silox)₃NbPMe

1-PMe₃

Figure 1. Molecular view of (silox)₃Ta(η-C₂H₃Et) (2-C₂H₃Et). Selected

bond distances (Å) and angles (deg) not in text: C1-C2, 1.395(7); C2-

C3, 1.494(7); C3-C4, 1.536(7); Si1-O, 1.664(2); Si2-O2, 1.678(3); Si3-

O3, 1.685(2); O1-Ta-C1, 91.4(2); O2-Ta-C1, 116.0(2); O3-Ta-C1,

111.8(2); O1-Ta-C2, 128.7(2); O2-Ta-C2, 95.6(2); O3-Ta-C2, 100.0-

(2); C1-Ta-C2, 37.9(2); C1-C2-Ta, 68.7(2); C2-C1-Ta, 73.3(3); C1-

C2-C3, 115.2(6); C2-C3-C4, 113.9(4); Ta-O1-Si1, 165.63(14); Ta-

O2-Si2, 167.89(15); Ta-O3-Si3, 172.81(16).

(silox)₃Nb(ole) + PMe₃

(4)

1-ole (ole ) H₂CCH^tBu, 90%; ^cC₅H₈, 50%)

Olefin isomerization proved not to be a problem in regard to

monitoring the olefin to alkylidene transformation. Thermolysis

of (silox)₃Nb(η-cis-C₂H₂Me₂) (1-c-C₂H₂Me₂) in benzene-d₆

afforded (silox)₃Nb(η-C₂H₃Et) (1-C₂H₃Et) after 3 d at 75 °C

(eq 5),

noted for slow cis-2-butene addition.⁴²The reactions were rapid

(<1 h) for all but norbornene, cyclopentene, and cyclohexene,

which were allowed to react for 12-24 h.⁴⁵Note that (silox)₃Ta-

C₆D₆

_{75 °C, 3 d}8

(silox)₃Nb(η-cis-C₂H₂Me₂)

1-c-C₂H₂Me₂

(silox) Nb(η-C H Et)

3 2 3

t

(η-C₂H₃Bu) (2-C₂H₃Bu), the neohexene derivative, could not

be cleanly synthesized, even when ∼30 equiv of H₂CdCH^tBu

were present. Use of ∼500 equiv H₂CdCH^tBu afforded a

1-C₂H₃Et

(5)

t

and the corresponding butene to butenylidene rearrangement

required more severe conditions (vide infra). Furthermore, no

evidence of disubstituted alkylidenes was observed, save those

derived from cyclic olefins. The niobium olefin adducts were

typically green, although the styrene derivatives tended toward

brown. ¹H and ¹³C{¹H} NMR spectral data on the olefin

complexes are given in Supporting Information.

substantial amount (>50%) of 2-C₂H₃Bu after 24 h at 23 °C,

t

but some alkylidene (silox)₃TadCHCH₂Bu (2dCHCH₂Bu) had

already formed, and additional species were present; hence the

rearrangement of this olefin was not pursued further. Spectral

data on the tantalum derivatives are also given in Supporting

Information.

Structure of (silox)₃Ta(η-C₂H₃Et) (2-C₂H₃Et). As a typical

asymmetric olefin complex, (silox)₃Ta(η-C₂H₃Et) (2-C₂H₃Et)

was selected for structural examination, and data collection and

refinement information may be found in Supporting Information.

Figure 1 reveals the molecular structure of 2-C₂H₃Et and gives

pertinent bond distances and angles. The C1-C2 double bond

has been lengthened to 1.395(7) Å, and the unit is oriented

primarily along the Ta-O1 vector, but slightly skewed such

that the Et group can best fit in the gap between the Si2 and

Si3 silox ligands. Given the noted reducing power of the

(silox)₃Ta (2) core, it is actually surprising that the C1-C2

lengthening is not greater;^46,47perhaps the steric bulk of the

silox groups is mitigating the potential back-donation of the

Ta(III) d²metal center, or its unique electronic features^38,39are

Synthesis of (silox)₃Ta(olefin) Complexes. Because of the

availability of (silox)₃Ta (2),⁴⁰the orange to red tantalum alkene

adducts were simply prepared (some previously)⁴²via the

addition of excess olefin (e.g., ∼1.1 equiv for styrene to ∼30

equiv for cyclohexene) in various hydrocarbon solvents accord-

ing to eq 6.

(silox) Ta

+ olefin (excess) f

3

2

(silox)₃Ta(ole)

2-ole (ole ) C₂H₄(62%); C₂H₃Me (56%);

(6)

C₂H₃Et (67%); ^cC₅H₈(63%); ^cC₆H₁₀(55%);

^cC₇H₁₀(47%); H₂C)CHC₆H₄-p-X,

X ) H (65%), OMe (62%), CF₃(76%)

(45) Hirsekorn, K. F.; Wolczanski, P. T.; Cundari, T. R., manuscript in

preparation.

A large excess of hindered olefins was used to minimize

competition from cyclometalation, which had been previously

(46) (a) Rietveld, M. H. P.; Teunissen, W.; Hagen, H.; van de Water, L.; Grove,

D. M.; van der Schaaf, P. A.; Muhlebach, A.; Kooijman, H.; Smeets, W.

J. J.; Veldman, N.; Spek, A. L.; van Koten, G. Organometallics 1997, 16,

1674-1684. (b) Abbenhuis, H. C. L.; Feiken, N.; Grove, D. M.; Jastrzebski,

J. T. B. H.; Kooijman, H.; Vandersluis, P.; Smeets, W. J. J.; Spek, S. L.;

van Koten, G. J. Am. Chem. Soc. 1992, 114, 9773-9781.

(47) (a) Schultz, A. S.; Brown, R. K.; Williams, J. M.; Schrock, R. R. J. Am.

Chem. Soc. 1981, 103, 169-176. (b) Burger, B.; Santarsiero, B. D.;

Trimmer, M. S.; Bercaw, J. E. J. Am. Chem. Soc. 1988, 110, 3134-3146.

(c) Visciglio, V. M.; Nguyen, M. T.; Clark, J. R.; Fanwick, P. E.; Rothwell,

I. P. Polyhedron 1996, 15, 551-554.

(41) Bonanno, J. B.; Henry, T. P.; Neithamer, D. R.; Wolczanski, P. T.;

Lobkovsky, E. B. J. Am. Chem. Soc. 1996, 118, 5132-5133.

(42) Covert, K. J.; Neithamer, D. R.; Zonnevylle, M. C.; LaPointe, R. E.;

Schaller, C. P.; Wolczanski, P. T. Inorg. Chem. 1991, 30, 2494-2508.

(43) Miller, R. L.; Toreki, R.; LaPointe, R. E.; Wolczanski, P. T.; Van Duyne,

G. D.; Roe, D. C. J. Am. Chem. Soc. 1993, 115, 5570-5588.

(44) Wolczanski, P. T. Polyhedron 1995, 14, 3335-3362.

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4811

A R T I C L E S

Scheme 1

Hirsekorn et al.

For the case of nexohexene, a mechanistic clue was uncovered

when another species was observed. Upon thermolysis at 103

t

°C for ∼12 h, (silox)₃Nb(η-C₂H₃Bu) (1-C₂H₃Bu) was converted

t

to the alkylidene (silox)₃NbdCHCH₂Bu (1dCHCH₂Bu, 25%),

but another species tentatively formulated as the “tuck-in”

neohexyl complex (silox)₂(^tBuCH₂CH₂)Nb(κ²-O,C-OSi^tBu₂-

t

CMe₂CH₂) (4-CH₂CH₂Bu, 25%) was also present. The disap-

pearance of 1-C₂H₃Bu corresponded roughly to the growth of

4-CH₂CH₂Bu, and initial rate studies provided a rate constant

t

of k ≈ 3.6 × 10^-7s^-1(∆G^q≈ 33.3 kcal/mol). Unfortunately,

the severity of overlapping resonances hampered integration

efforts to the point where simulation of the concentration

vs time profiles of the three species that were obtained did

not lead to convergence. In time, complete conversion to

t

1dCHCH₂Bu was noted. It is tempting to conclude that a

t

1-C₂H₃Bu h 4-CH₂CH₂Bu h 1dCHCH₂Bu sequence is

operable, but the possibility that cyclometalation⁵⁰is simply a

side reaction cannot be discounted.

2. Kinetics and Thermodynamics. The data affiliated with

each case are given in Table 1, along with pertinent activation

parameters obtained from Eyring plots that typically were

limited to a temperature range of ∼40 °C because of the elevated

temperatures required for reasonable rates. Aside from ethylene

and styrene, no K_alk/olevalues were obtained because only the

alkylidene was observed after a period of time. Since the

thermolyses were conducted at relatively high temperatures, a

common temperature was sought as a reference state at which

the niobium and tantalum activation free energies could be

compared. Although 103 °C was a rather low temperature for

the observation of most niobium reactions, it was convenient

for the equivalent, more elaborate tantalum processes, and

therefore served as an appropriate reference condition.

hampering binding. The d(Ta-O1) of 1.943(2) Å is longer

than the remaining siloxide lengths (d(Ta-O2) ) 1.907(2) Å,

d(Ta-O3) ) 1.892(2) Å) in response to the subtle influence

of the 1-butene binding, and the O2-Ta-O3 angle has opened

up (119.25(11)°) relative to O1-Ta-O2 (104.68(10)°) and

O1-Ta-O3 (109.50(10)°) to accommodate the ethyl group. As

expected, binding of the 1-butene is asymmetric with the

â-carbon-tantalum distance ∼0.06 Å longer than the R-carbon

(2.174(4) vs 2.115(4) Å). The nearest silox-hydrogen to the

â-carbon is 3.30 Å away, and its carbon is 4.05 Å distant.

For the cases listed in Table 1, no intermediates were detected,

and the rate of olefin complex rearrangement is styrene ≈

ethylene < para-methoxystyrene < propene < 1-butene <

cyclopentene < norbornene < cyclohexene. As the size of the

olefin increases, and as more electron-donating substituents are

present, the rate of rearrangement becomes faster. For the

styrenes, the electron-withdrawing nature of the phenyl groups

must compensate for the larger size of this olefin, rendering

these cases quite slow. Note that having an electron donating

substituent (OMe) in the para-position speeds up the rate of

rearrangement relative to styrene, as expected.

Niobium Olefin to Alkylidene. 1. Observations. Thermoly-

ses of the niobium olefin complexes (silox)₃Nb(ole) (ole )

C₂H₄, 1-C₂H₄; C₂H₃Me, 1-C₂H₃Me; C₂H₃Et, 1-C₂H₃Et;^{36 c}C₅H₈,

c

1-^cC₅H₈; C₆H₁₀, 1-^cC₆H₁₀;^{36 c}C₇H₁₀, 1-^cC₇H₁₀; H₂CCHC₆H₄-

p-X (X ) H, 1-C₂H₃Ph; OMe, 1-C₂H₃Ph-p-OMe)) were

undertaken at various temperatures. In these cases, smooth and

reversible formation of the respective alkylidene complexes

(silox)₃Nb(alk) (alk ) CHMe, 1dCHMe; CHEt, 1dCHEt;

CHⁿPr, 1dCHⁿPr; ^cC₅H₈, 1d^cC₅H₈; ^cC₆H₁₀, 1d^cC₆H₁₀; ^cC₇H₁₀,

1d^cC₇H₁₀; CHCH₂C₆H₄-p-X (X ) H, 1dCHCH₂Ph; OMe, 1d

CHCH₂Ph-p-OMe)) was observed in benzene-d₆according to

In the niobium ethylene case, a van’t Hoff plot corroborated

the activation energies and revealed a large positive ∆S° ) 12.5

(10) eu that helped compensate for a ∆H° ) 6.3 (4) kcal/mol.

With only one carbon bound to Nb, (silox)₃NbdCHMe (1d

CHMe) is less constrained in both the alkylidene (e.g., methyl

rotor, etc.) and siloxide periphery, rendering it entropically more

favorable than the parent ethylene species, (silox)₃Nb(η²-C₂H₄)

(1-C₂H₄). It is suspected that the alkylidene is more favored

entropically in all cases, although perhaps less so for the cyclic

derivatives.

1

Scheme 1. The compounds were assayed by H and ¹³C{¹H}

NMR spectroscopy, but difficulties were encountered in observ-

ing the alkylidene carbons because of quadrupolar broadening

by niobium. Heteronuclear quantum coherence methods permit-

ted assessment of the chemical shifts of 1dCHEt (δ 228) and

1dCHⁿPr (δ 249.0), while synthesis of (silox)₃Nbd¹³CH¹³CH₃

(1d¹³CH¹³CH₃, δ 213.0, J_CH) 114 Hz) allowed direct detection

of its alkylidene resonance and evidence of a modest agostic

interaction. Since J_CHfor niobium alkylidenes cannot be easily

observed, IR spectra were carefully scrutinized, but absorbances

characteristic of agostic interactions were not noted.^23,48,49NMR

spectral data are given in Supporting Information.

3. Attempts at r-Substituent Effects. In an attempt at further

electronic substituent studies, the 170 °C thermolysis of

(48) Brookhart, M.; Green, M. L. H.; Wong, L. Prog. Inorg. Chem. 1988, 36,

1-124.

(49) Goddard, R. J.; Hoffmann, R.; Jemmis, E. D. J. Am. Chem. Soc. 1980,

102, 7667-7676.

(50) (a) Rothwell, I. P. Acc. Chem. Res. 1988, 21, 153-159. (b) Rothwell, I. P.

Polyhedron 1985, 4, 177-200.

9

4812 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

A R T I C L E S

Table 1. Rate Constants^a(k_f(Scheme 1) unless Noted)^band Activation Parameters for the (silox)₃Nb(ol) (1-ol) h (silox)₃Nb(alk) (1dalk)

Rearrangement (∆G°, ∆G^q, and ∆H^qin kcal/mol; ∆S^qin eu) and Related Processes

1

q

reaction

T (°

C)^c

k (

×

10⁵s^-

)

∆

G_f^q

G₁₀₃

∆

H^q

∆

S^q

∆G₁₀₃°

°C

°

C

1-C₂H₄h 1dCHMe

155

168

180

196

0.371(1)

1.22(2)^b

1.14(2)

35.5(3)

31.9(3)

-9.6(2)

1.60(44)^d

2.2(10)^e

33.3(8)^b

24.7(6)^b

-22.8(14)^b

2.75(5)^b

3.20(6)

6.88(8)^b

10.63(1)

17.54(4)^b

1.63(2)

3.54(2)

10.3(3)

1-C₂H₃Me h 1dCHEt

1-C₂H₃Et h 1dCHⁿPr

140

155

166

180

150

166

33.4(41)

32.7(2)

29.6(40)

28.3(2)

-10.1(25)

-11.6(3)

42.2(19)

6.44(6)

22.7(5)

22.9(3)^f

72.0(16)

130(1)

33.3(1)^f

182

190

166^f

103

1-D₂CCHEt f 1dCDⁿPr-d₂

9.24(7)^f

34.1(1)^f

33.3^g

t

1-C₂H₃Bu f 4-CH₂CH₂Bu

1-^cC₅H₈h 1dC(CH₂)₃CH₂

1-^cC₇H₁₀h 1d^cC₇H₁₀

0.036^g

0.385(6)

0.765(11)

0.00404^h

111(6)ⁱ

31.50(6)

30.98(6)

34.9^h

166

32.0(1)ⁱ

36.5(1)^h

32.7(1)ⁱ

0.617(17)^h

49.8(8)ⁱ

0.0787(8)

0.225(4)

1.18(3)

1-^cC₇H₈D₂h 1d^cC₇H₉D-d₂

1-^cC₆H₁₀h 1d ^cC₆H₁₀

166

60

70

29.5(2)

23.8(2)

-15.2(3)

86

103

155

5.38(17)

0.321(2)

0.325(5)^b

0.566(4)

0.337(10)^b

1--C₂H₃Ph h 1dCHCH₂Ph

36.11(3)

36.11(6)

35.63(4)

36.1(1)

35.5 (est)^j

34.9 (est)^b,l

35.0 (est)^j

34.8 (est)^b,l

0.6^k

0.6 (est)^e,l

0.2^k

1-C₂H₃Ph-p-OMe h 1dCHCH₂Ph-p-OMe

155

0.2 (est) ^e,l

^aRate constants were obtained from first-order loss of 1-ole, approach to equilibrium kinetics, or from simulation of the latter as necessary and are the

average of three simultaneous kinetics runs. ^bRate constants and parameters corresponding to k_r. ^cTemperatures are estimated to be (0.5 °C. ^dFrom van’t

Hoff plot of direct measurements of K_alk/olat the temperatures indicated: ∆H° ) 6.30(44) kcal/mol and ∆S° ) 12.5(10) eu. ^eTaken from ∆G_f^q- ∆G_r^q.

^fConducted in tandem (triplicate tubes for each) to afford k_H/k_D) z_f) 2.5 at 166 °C. ^gRough rate constant obtained from initial rates on the disappearance

t

of 1-C₂H₃Bu. ^hRate constant for k_ras determined from KIE experiments at 166 °C as described in text. The ∆G_r^qat 103 °C was obtained by assuming ∆S°

≈ 12 eu. ⁱConducted in tandem (triplicate tubes for each) to afford k_H/k_D) z_f) 2.3 at 166 °C. ^jEstimated using ∆S^q) -12 eu, a reasonable average of

k

similar activation entropies. K_alk/olmeasured directly at 155 °C and corrected to 103 °C using ∆S° ≈ 13 from the ethylene case. ^lEstimated using ∆S^q≈

-24 eu obtained from an estimate of -12 eu for the forward step and ∆S° ≈ 12 eu from the ethylene case.

(silox)₃Nb(η-C₂H₃C₆H₄-p-CF₃) (1-C₂H₃Ph-p-CF₃) was under-

taken, but CF bond activation^51-55was noted instead of the

expected alkylidene. To examine the potential for Fischer-type

carbenes to rearrange, (silox)₃NbPMe₃(1-PMe₃) was exposed

to methylvinyl ether and vinylfluoride, respectively. It was hoped

that (silox)₃Nb(η-C₂H₃X) (X ) OMe, 1-C₂H₃OMe; F, 1-C₂H₃F)

would form and be converted to (silox)₃NbdCXCH₃or (silox)₃-

NbdCHCH₂X over time, but NMR tube experiments indicated

that C-X bond activation occurred to give (silox)₃(H₂CdCH)-

NbX (X ) OMe, 5-OMe; F, 5-F) (eq 7).

dihydrofuran to (silox)₃Ta (2),⁵⁶these results were anticipated,

but it was nonetheless disappointing that formation of Fischer

carbenes seemed untenable.

Structure of (silox)₃Nbd^cC₆H₁₀(1d^cC₆H₁₀). The niobium

cyclohexylidene complex, (silox)₃Nbd^cC₆H₁₀(1d^cC₆H₁₀), was

chosen for further structural characterization (see Supporting

Information for crystallographic details). One of four crystal-

lographically distinct molecules of 1d^cC₆H₁₀is shown in Figure

2, and some average bond distances and angles are also listed.

The average d(NbdC) is 1.956(18) Å, which is typical for a

niobium alkylidene.^{22,23,36,57,58}While no significant deviations

in average Nb-O bond distances (1.917(28) Å) are observed,

one O-Nb-O angle (125.3(15)° ave) was substantially wider

than the remaining O-Nb-O angles of 107.7(13)° (ave), despite

similar O-Nb-C angles (105.1(21)° ave). The NbC(C_R)₂plane

is nearly perpendicular to one Nb-O bond, and the chair of

benzene-d₆

(silox)₃NbPMe₃+ H₂CdCHX

1-PMe₃

8

23 °C

(silox)₃(H₂CdCH)NbX + PMe₃

X ) OMe, 5-OMe; F, 5-F

(7)

Given the precedent set by the oxidative addition of 2,3-

(55) (a) Ferrando-Miguel, G.; Gerard, H.; Eisenstein, O.; Caulton, K. G. Inorg.

Chem. 2002, 41, 644-6449. (b) Huang, D. J.; Koren, P. R.; Folting, K.;

Davidson, E. R.; Caulton, K. G. J. Am. Chem. Soc. 2000, 122, 8916-

8931.

(56) Bonanno, J. B.; Henry, T. P.; Neithamer, D. R.; Wolczanski, P. T.;

Lobkovsky, E. B. J. Am. Chem. Soc. 1996, 118, 5132-5133.

(57) Caselli, A.; Solari, E.; Scopelliti, R.; Floriani, C. J. Am. Chem. Soc. 1999,

121, 8296-8305.

(58) Nugent, W. A.; Mayer, J. M. Metal-Ligand Multiple Bonds; Wiley-

Interscience: New York, 1988.

(51) (a) Clot, E.; Megret, C.; Kraft, B. M.; Eisenstein, O.; Jones, W. D. J. Am.

Chem. Soc. 2004, 126, 5647-5653. (b) Jones, W. D. J. Chem. Soc., Dalton

Trans. 2003, 3991-3995.

(52) Strazisar, S. A.; Wolczanski, P. T. J. Am. Chem. Soc. 2001, 123, 4728-

4740.

(53) Watson, L. A.; Yandulov, D. V.; Caulton, K. G. J. Am. Chem. Soc. 2001,

123, 603-611.

(54) Edelbach, B. L.; Jones, W. D. J. Am. Chem. Soc. 1997, 119, 7734-7742.

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4813

A R T I C L E S

Hirsekorn et al.

Scheme 2

Figure 2. Molecular view of one of the four crystallographically distinct

molecules of (silox)₃Nbd^cC₆H₁₀(1d^cC₆H₁₀). Selected bond distances (Å)

and angles (deg) not in text: Si-O(ave), 1.660(28); C-C(ring ave), 1.519-

(47); Nb-C-C(ave), 124.5(15); C-C-C(ring ave), 110.9(22); Nb-O-

Si(ave), 169.8(65).

the cyclohexane ring is disposed toward the unique, wide

O-Nb-O angle, which presumably reflects its steric influence.

The closest silox-hydrogen to the alkylidene carbon is a mere

3.06 Å away, and its carbon is 4.00 Å distant.

Tantalum Olefin to Alkylidene. 1. Observations. Ther-

molyses of the tantalum olefin complexes (silox)₃Ta(ole) (ole

) C₂H₄, 2-C₂H₄; C₂H₃Me, 2-C₂H₃Me; C₂H₃Et, 2-C₂H₃Et;

Rate constants coupled to species of low concentration were

intrinsically less precise and probably less accurate, simply

because of the quality of integration. Entropy of activation

values, critical for estimating ∆G^q’s at 103 °C in many cases,

was estimated using the most accurate rate and thermodynamic

(e.g., ∆S°) data available. The following ∆S^qvalues were

c

^cC₅H₈, 2-^cC₅H₈; C₆H₁₀, 2-^cC₆H₁₀; H₂CCHC₆H₄-p-X (X ) H,

2-C₂H₃Ph; OMe, 2-C₂H₃Ph-p-OMe; CF₃, 2-C₂H₃Ph-p-CF₃))

were undertaken at temperatures somewhat lower than those

of the niobium study. As Scheme 2 reveals, typically two

species were observed: the “tuck-in” alkyls (silox)₂(R)Ta-

(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-R: R ) Et, 6-Et; ⁿPr, 6-ⁿPr; ⁿBu,

q

deemed reasonable estimates for several cases: ∆S₁≈ -10

c

6-ⁿBu; Pe, 6-^cPe; Hex, 6-^cHex; CH₂CH₂C₆H₄-p-X (X ) H,

6-CH₂CH₂Ph; OMe, 6-CH₂CH₂Ph-p-OMe; F, 6-CH₂CH₂-

Ph-p-F)), and the alkylidenes (silox)₃Ta(alk) (alk ) CHMe,

2dCHMe; CHEt, 2dCHEt; CHⁿPr, 2dCHⁿPr; ^cC₅H₈, 2d^cC₅H₈;

^cC₆H₁₀, 2d^cC₆H₁₀; CHCH₂C₆H₄-p-X (X ) H, 2dCHCH₂Ph;

OMe, 2dCHCH₂Ph-p-OMe; F, 2dCHCH₂Ph-p-F)). Only the

norbornene derivative, (silox)₃Ta(η-^cC₇H₁₀) (2-^cC₇H₁₀), smoothly

converted to the norbornylidene (silox)₃Tad^cC₇H₁₀(2d^cC₇H₁₀)

without observation of the “tuck-in” alkyl (silox)₂(2-norbornyl)-

Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-^cC₇H₁₁). For the tantalum

alkylidenes, ¹³C resonances of the alkylidenes were easily

obtained, and their average chemical shift was 228.5(33) ppm,

with exceptions being 2dCHMe (δ 204.71, J_CH) 111 Hz) and

2dCHⁿPr (δ 254.84, J_CH) 108 Hz); again, only modest agostic

interactions were evident.^23,48,49

q

eu, ∆S_-1≈ ∆S₂≈ -7 eu, and ∆S_-2≈ -27 eu. With the aid

of ∆S° values determined from equilibrium measurements, the

q

styrene ∆G_103°Cvalues were estimated somewhat differently,

q

with ∆S₁≈ -8 eu, ∆S_-1≈ -14 eu, ∆S₂≈ -8 eu, and ∆S_-2

≈ -27 eu. Details are provided as Supporting Information.

As a partial check of the ∆S^qestimates, the temperature

dependence of the rate of butene complex conversion to its

“tuck-in” butyl, (silox)₃Ta(η-C₂H₃Et) (2-C₂H₃Et) f (silox)₂-

(ⁿPrCH₂)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-ⁿBu), was measured

from 78 to 120 °C. The Eyring plot yielded activation parameters

of ∆H₁) 24.8(1) kcal/mol and ∆S^q) -11(1) eu, in support

q

of the above estimates. Note that the entropy of activation is

similar to that found for the cyclometalation of (silox)₃Ta (2)

to (silox)₂HTa(κ²-O,C-OSi^tBu₂CMe₂CH₂) (8, eq 8).

benzene-d₆

2. Simulations and Assessment. Table 2 lists the cases

studied and reveals the first-order rate constants assigned to the

steps in Scheme 2. In many casessethylene, cyclopentene, and

the styrenessall three species could be observed throughout

the course of reaction, but in somespropene, 1-butene, and

cyclohexenesthe intermediate “tuck-in” alkyls (6-R) and alkyl-

idenes (2dalk) proved to be sufficiently stable relative to the

starting olefin complexes that no 2-ole was observable at

equilibrium. Simulation of the concentration vs time profiles

2

t

(silox)₃Ta

8

(silox)₂HTa(κ -O,C-OSi Bu₂CMe₂CH₂)

2

8

(8)

The temperature dependence of this process yielded a ∆H^qof

19.9(27) kcal/mol and ∆S^qof -15(4) eu over a 52-104 °C

range.⁴²

3. Kinetics and Thermodynamics. Since the rate constants

for all steps in Scheme 2 can be measured or estimated, a

thermodynamic assessment of the rearrangement can be made.

The alkylidene complexes are favored over the olefins in all

1

obtained via H NMR spectra was required to determine all

possible rate constants in each case.

9

4814 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

a

A R T I C L E S

Table 2. Rate Constants (×10⁴s^-1

)

and Activation Free Energies (∆G^qin kcal/mol) for the (silox)₃Ta(ol) (2-ol) h

(silox)₂(R)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-R) h (silox)₃Ta(alk) (2dalk) Rearrangement (Scheme 2)

reaction

T (

°

C)^b

k₁

k_-

1

k₂

k_-

2

∆ ∆ ) ∆ ∆ )

G₁₀₃^q(k₁) G₁₀₃^q(k_-G₁₀₃^q(k₂) G₁₀₃^q(k_-

°C °C 1 °C °C 2

2-C₂H₄h 6-Et h 2dCHMe

130.5 15.3(4)

140.5 26.9(3)^e

5.90(8)

0.177(11)

1.90(21) 28.4(39)^c29.3(23)^c31.9(27)^c29.5(35)^c,d

3.67(5)^e

10.97(9)^e0.472(7)^e

150

33.4(19)

14.2(8)

68.5(50)

0.683(33)

2.70(14)

3.91(24) 29.1(3)^f

9.90(12)

30.0(3)^f

32.7(3)^f30.2(3)^f

170.5 146(12)

2-C₂H₃Me f 6-ⁿPr h 2dCHEt

2-C₂H₃Et f 6-ⁿBu h 2dCHⁿPr

103

155

78

1.304(7)

28.9(1)

32.5^g

30.4^g

1.45(2)

1.19(3)

5.07(20)

0.103(8)^h

0.432(4)^h

1.42(4)^h

4.87(5)^h

93

103

120

155

103

166

103

28.8(1)

32.6ⁱ

30.6ⁱ

3.88(4)

28.8(1)^j

29.1(1)^j

2-C₂H₃Et f 6-ⁿBu^j

1.40(6)^j

n

2-D₂CCHEt f 6-CD₂Pr^j

0.924(7)^j

6-ⁿBu h 2dCHⁿPr^k,l

3.38(8)^k

32.6^m

34.2ⁿ

n

k

6-CD₂Pr f 2dCDⁿPr-d₂

0.547(6)^k

0.387(10)^l

3.77(12)

l

6-CHDⁿPr f 2dCHⁿPr-d₂

34.5°

28.1(9)

2-^cC₅H₈h 6-^cPe h 2dC(CH₂)₃CH₂

2-^cC₇H₁₀h 2d^cC₇H₁₀

2.97(5)

2.16(6)

1.01(3)

1.90(9)

28.2(2)

28.5(2)

29.1(9)

28.6(1)

0.584(5)^p

0.0589(46)^q

5.54(6)

29.5(1)^p

31.2(2)^q

2-^cC₆H₁₀f 6-^cHex h 2d^cC₆H₁₀

103

4.14(16)

27.8(1)

28.0(1)

2-C₂H₃Ph h 6-CH₂CH₂Ph h 2dCHCH₂Ph^s140

0.267(14)

1.02(2)^t

2.31(4)

7.62(27)

1.83(2)

0.0183(9) 0.0820(24) 0.350(28) 32.9(7)^r

35.4(43)^r34.5(20)^r32.6(17)^r

155

166

180

0.178(5)^t0.57(2)^t

0.377(11) 1.40(3)

1.92(7)^t

4.08(9)

32.8(1)^u,V33.9(8)^u,V33.3(4)^u,V31.3(4)^u,V

1.28(5)

0.191(7) 0.70(4)

6.97(5)

13.1(10)

2.48(9)

2-C₂H₃Ph-p-OMe h 6-CH₂CH₂Ph-p-OMe h 155

2dCHCH₂Ph-p-OMe

2-C₂H₃Ph-p-CF₃h 6-CH₂CH₂Ph-p-CF₃h

2dCHCH₂Ph-p-CF₃

32.3^u,w

33.9^u,w

34.2^u,x

33.1^u,w

33.7^u,x

31.0^u,w

31.8^u,x

155

0.256(4)

0.136(2) 0.336(6)

0.984(6) 33.9^u,x

(silox)₃Ta (2) f

52

56

64

71

75

0.572(6)^y

1.0(1)^z

(silox)₂HTa(OSi^tBu₂CMe₂CH₂) (8)

2.1(1)^z

4.17(4)^y

4.8(1)^z

84

104

11.9(2)^y

49.9(13)^y

26.1(1)

^aRate constants were obtained from simulation of the concentration vs T profile of the three species and are the average of three simultaneous kinetics

runs unless otherwise noted; ∆G^q’s were obtained directly from k’s at 103 °C or from Eyring plots or estimates as noted. ^bTemperatures are estimated to

be (0.5 °C. ^cActivation free energies determined from Eyring plots (∆H^qin kcal/mol, ∆S^qin eu): ∆H₁^q) 18.9(27), ∆S₁^q) -25.3(73); ∆H_-1^q) 20.7(1),

∆S_-1^q) -22.9(60); ∆H₂^q) 22.6(19), ∆S₂^q) -25(5); ∆H_-2^q) 13(2), ∆S_-2^q) -43(57). ^dThe error in ∆S_-2^qwas so great that this value was determined

q

from measurement of K(2) and thus ∆G°(2) (∆G_-2) ∆G°(2) - ∆G₂). From van’t Hoff plot of K(1) ) k₁/k_-1: ∆H°(1) ) -1.7(1) kcal/mol, ∆S°(1) )

-2.4(1) eu. From van’t Hoff plot of K(2) ) k₂/k_-2: ∆H°(2) ) 9.5(6) kcal/mol, ∆S°(2) ) 19.0(14) eu. ^eTaken from the average of two simultaneous kinetics

q

runs. ^fEstimated from the 130.5 - 170.5 °C data using ∆S₁≈ -10 eu, ∆S_-1≈ ∆S₂≈ -7 eu, and ∆S_-2) -27 eu; the ∆G_103°C^q’s for each step were

calculated from each temperature and averaged. ^gEstimated from the 155 °C data (∆G₂^q) 32.9(1) and ∆G_-2^q) 31.8(1) kcal/mol) using ∆S₂≈ -7 eu and

q

∆S_-2≈ -27 eu. ^hFrom an Eyring plot, the activation parameters for k₁are ∆H₁^q) 24.8 (1) and ∆S₁^q) -11(1). ⁱEstimated from the 155 °C data (∆G₂

q

) 33.0(1) and ∆G_-2) 32.0(1) kcal/mol) using ∆S₂≈ -7 eu and ∆S_-2≈ -27 eu. ^jFrom a tandem measurement, 1.52 ) z₁′²) (1.23)²as in Scheme

q

4. ^kFrom fits of the approach to equilibria via tandem measurements, z₂z₂′ ) 6.2 as in Scheme 4. ^lFrom fits of the approach to equilibria via tandem

q

measurements, z₂) 4.4 as in Scheme 4. By difference with the previous case, z₂′ ) 1.4. ^mCalculated from ∆G_166°C) 33.0(1) assuming ∆S₂) -7 eu.

q

ⁿCalculated from ∆G_166°C) 34.6(1) assuming ∆S₂) -7 eu. ^oCalculated from ∆G_166°C) 34.9(1) assuming ∆S₂) -7 eu. ^pCorresponds to k_fof

Scheme 2. ^qCorresponds to k_rof Scheme 2. ^rActivation free energies determined from Eyring plots (∆H^qin kcal/mol, ∆S^qin eu): ∆H^q(k₁) ) 30.0(7),

∆S^q(k₁) ) -7.5(3); ∆H^q(k_-1) ) 37.9(43), ∆S^q(k_-1) ) 6.64(12); ∆H^q(k₂)^q) 39.7(20), ∆S^q(k₂) ) 13.8(10); ∆H^q(k_-2) ) 31.7(17), ∆S^q(k_-2) ) -2.5(2).

^sFrom van’t Hoff plots using K(1)_140°C) 5.10, K(1)_155°C) 5.41, K(1)_166°C) 5.58, K(1)_180°C) 5.90, K(2)_140°C) 0.19, K(2)_155°C) 0.27, K(2)_166°C) 0.34,

and K(2)_180°C) 0.51: ∆H°(1) ) 1.33(10) kcal/mol, ∆S°(1) ) 6.46(22) eu, ∆H°(2) ) 9.05(24) kcal/mol, ∆S°(2) ) 18.56(38) eu. ^tAt 155 °C (kcal/mol):

∆G₁) 33.2(1), ∆G_-1) 34.7, ∆G₂) 33.7, ∆G_-2) 32.6. ^uEstimated from the 140-180 °C data using ∆S₁≈ -8 eu, ∆S_-1≈ -14 eu, ∆S₂) -8

q

eu, and ∆S_-2^q) -27 eu. ^VThe ∆G_103°C^q’s for each step were calculated at each temperature and averaged. ^wAt 155 °C (kcal/mol): ∆G₁^q) 32.7(1), ∆G_-1

q

) 34.6(1), ∆G₂) 33.5(1), ∆G_-2) 32.4(1). ^xAt 155 °C (kcal/mol): ∆G₁) 34.4(1), ∆G_-1) 34.9(1), ∆G₂) 34.1(1), ∆G_-2) 33.2(1). ^yRate

constants for cyclometalation were determined by ¹H NMR spectroscopy. ^zRate constants were determined from UV-vis spectroscopy according to ref 42.

From an Eyring plot of all rate constants, the cyclometalation of 2 to 8 affords ∆H^q) 19.9(27) kcal/mol and ∆S^q) -15(4) eu; at 103 °C, ∆G^q) 25.5(30).

cases except ethylene and the styrenes, just as in the niobium

system. However, the tuck-in alkyl complexes are the most

stable of the three species, except for the norbornene, cyclo-

pentene, and cyclohexene cases. In the latter two, the alkylidenes

are slightly favored over the tuck-in alkyls.

While the rate of the niobium rearrangements correlates

inversely with the expected stabilities of the olefin complexes,

the tantalum results are less easy to interpret. Since the tuck-in

alkyl intermediates are observed, consideration of a rate-

determining step in going from olefin complex to alkylidene is

no longer appropriate. In many cases, the first transition state

has the highest free energy, yet k₂< k₁because of the stability

of the tuck-in alkyl. In general, the tantalum rates are substan-

tially faster than those of niobium, and the same rough trends

exist, at least in the context of the olefin to alkylidene

transformation. The styrenes and ethylene are the slowest, the

propene and 1-butene are next, and the speedy rearrangements

are the cyclics and norbornene.

Structure and Dynamics of (silox)₃BuTa(K²-O,C-OSi^tBu₂-

n

CMe₂CH₂) (6-ⁿBu). X-ray diffraction quality crystals of the

n

tantalum tuck-in butyl derivative, (silox)₃BuTa(κ²-O,C-OSi^t-

Bu₂CMe₂CH₂) (6-ⁿBu), were obtained, and its molecular

structure is given in Figure 3 along with pertinent bond distances

and angles; further structural information is given in Supporting

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4815

A R T I C L E S

Hirsekorn et al.

in comparison to normal silox geometric parameters. Despite

this strain, the d(Ta-C2) of 2.211(3) Å is only slightly longer

than the comparative butyl distance, d(Ta-C13) ) 2.178(4) Å.

The remaining axial/equatorial angles support the distorted tbp

geometry and are consistent with intersilox steric repulsions

being the most influential about the core.

The spectral signatures of the tuck-in alkyl complexes,

(silox)₃(R)M(κ²-O,C-OSi^tBu₂CMe₂CH₂) (M ) Nb, R ) CH₂-

t

CH₂Bu, 4-CH₂CH₂Bu; Ta, 6-R), indicate mirror symmetry,

n

while the solid-state structure of (silox)₃BuTa(κ²-O,C-OSi^tBu₂-

CMe₂CH₂) (6-ⁿBu) is asymmetric. Variable temperature ¹H

NMR spectroscopy on (silox)₃PrTa(κ²-O,C-OSi^tBu₂CMe₂CH₂)

n

(6-ⁿPr), chosen for ease of analysis, revealed three coalescence

phenomena associated with the R-CH₂(k ) 1500 s^-1at T_c)

218 K, ∆G^q) 10.4(5) kcal/mol) and â-CH₂(k ) 730 s^-1at T_c

) 221 K, ∆G^q) 9.9(5) kcal/mol) methylenes of the ⁿPr group,

and the methylene (k ) 300 s^-1at T_c) 223 K, ∆G^q) 9.5(5)

kcal/mol) of the cyclometalation arm. The activation free

energies affiliated with these resonances are consistent with the

relatively low barriers common to axial and equatorial exchange

mechanisms in five-coordination.⁵⁹

n

Figure 3. Molecular view of (silox)₃BuTa(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-

ⁿBu). Selected bond distances (Å) and angles (deg) not in text: C1-C2,

1.556(5); C13-C14, 1.520(5); C14-C15, 1.511(6); C15-C16, 1.521(7);

Si1-O1, 1.676(3); Si2-O2, 1.674(2); Si3-O3, 1.682(2); O1-Ta-C13,

86.07(12); O2-Ta-C13, 88.48(12); Ta-C13-C14, 115.5(3); C13-C14-

C15, 114.8(4); C14-C15-C16, 112.9(4); Ta-O2-Si2, 171.27(15); Ta-

O3-Si3, 174.98(16).

Olefin to Alkylidene Mechanism. 1. Mechanistic Paths.

The previous discussion regarding the rates and thermodynamics

of the olefin to tuck-in alkyl to alkylidene rearrangement was

predicated on the tuck-in alkyl as a viable intermediate. This

need not be the case. As Scheme 3 shows, in some scenarios

the tuck-in alkyl would be a side process of no consequence to

the reaction of interest. For example, the olefin complex could

rearrange via a one-step 1,2-H migration. Likewise, a vinylic

CH bond activation, followed by a 1,3-H migration leads to

the same product if the R-positions of the olefin were labeled.

Since these processes are not readily distinguished, either

constitutes an H-transfer path.

Information. 6-ⁿBu is best described as a distorted trigonal

bipyramid, with the less electronegative alkyl substituents,

cyclometalated C2 and Bu C13, and silox O3 occupying the

n

equatorial plane (/O3-Ta-C2 ) 121.06(13)°, /O3-Ta-C13

) 111.41(13)°, and /C2-Ta-C13 ) 127.10(15)°).⁵²The

remaining silox groups are pseudoaxial, with /O1-Ta-O2 )

154.06(10°), and d(Ta-O1) ) 1.937(2) Å and d(Ta-O2) )

1.920(2) Å, which are longer than the equatorial d(Ta-O3) of

1.878(2) Å, as expected. The bite angle of the cyclometalated

unit (/O1-Ta-C2 ) 78.84(12)°) is less than 90°, and the

remaining angles of the five-membered ring manifest significant

strain (/Ta-O1-Si1 ) 128.34(14)°, /O1-Si1-C1 ) 97.97-

(14)°, /Si1-C1-C2 ) 101.9(2)°, /Ta-C2-C1 ) 118.8(2)°)

A second possibility arises from loss of olefin and cyclo-

metalation of (silox)₃M (M ) Nb, 1; Ta, 2) to the original tuck-

in species, the hydride (silox)₂HM(κ²-O,C-OSi^tBu₂CMe₂CH₂)

(M ) Nb, 7;³⁸Ta, 8).⁴²Reinsertion of the olefin to afford the

Scheme 3

9

4816 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

A R T I C L E S

benzene-d₆

tuck-in alkyl, followed by an abstraction by the cyclometalation

arm on the R-carbon of the primary alkyl, leads to the alkylidene.

This path is termed the olefin loss mechanism.

(silox)₃Nb(η-C₂H₃Me) + C₂H

1-C₂H₃Me

₄y

z

155 °C

(silox)₃Nb(η-C₂H₄) + C₂H₃Me

1-C₂H₄

(11)

The third possibility is the one implied previously. A

δ-abstraction of a hydrogen from a silox CH bond by the

â-carbon of the olefin can lead directly to the “tuck-in” alkyl.

A subsequent R-abstraction by the cyclometalation arm on the

primary alkyl provides the alkylidene. Consideration of this

double abstraction path leads to another possibility. If a

monosubstituted olefin were able to abstract hydrogen from a

silox CH bond utilizing the unsubstituted R-carbon, a disubsti-

tuted alkylidene could result, but none has been directly

obserVed for monosubstituted olefin substrates. More impor-

tantly, if δ-abstraction led to an intermediate tuck-in secondary

alkyl, reversibility of this reactionsa â-abstraction by the

cyclometalation arm on the alkylswould eventually lead to loss

of label from the R-positions of a monosubstituted olefin to the

silox groups, which is referred to as scrambling in the scheme.

The 1-C₂H₃Me + C₂H₄experiment is consistent with the

olefin loss pathway, provided either the insertion or the

R-abstraction steps are rate-determining, since ethylene would

be expected to trap (silox)₃Nb (1), precursor to the tuck-in

(silox)₂HNb(κ²-O,C-OSi^tBu₂CMe₂CH₂) (7), faster than propene

would rebind. In contrast, if the same criteria were applied to

the ¹³C₂H₄experiment, loss of labeled ethylene from the starting

complex should be greater than observed; thus the olefin loss

path was considered unlikely. No order in cyclohexene was

observed when the rearrangement of (silox)₃Nb(η-^cC₆H₁₀) (2-

^cC₆H₁₀) was conducted in the presence of excess (19 equiv)

olefin.

Thermolysis (103 °C, ∼4 h) of cyclopentene adduct (silox)₃Ta-

(η-^cC₅H₈) (2-^cC₅H₈) with ∼15 equiv of ethylene present afforded

only the two rearrangement products: tuck-in cyclopentyl

(silox)₂(^cPe)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-Pe, 36%) and cy-

clopentylidene (silox)₃Tad^cC₅H₈(2d^cC₅H₈, 48%, eq 12).

2. Elucidation. Previous experience with the cyclometala-

tion of (silox)₃M (M ) Nb, 1; Ta, 2) to (silox)₂HM(κ²-O,C-

OSi^tBu₂CMe₂CH₂) (M ) Nb, 7;³⁸Ta, 8)⁴²pointed toward the

olefin loss path as a distinct possibility; hence various labeling

experiments were devised to assess it. First, the perdeuterated

ethylene complex, (silox)₃Nb(η-C₂D₄) (1-C₂D₄), was thermo-

lyzed at 155 °C to probe for deuterium incorporation into the

silox ligands (eq 9).

(silox)₃Ta(η-^cC₅H₈) + C₂H

2-^cC₅H₈

₄h 6-^cPe + C H h

2

4

(silox)₃Tad^cC₅H₈+ C₂H₄

2d^cC₅H₈

(12)

benzene-d₆

(silox) Nb(η-C D )

y

z

3

2

4

155 °C

It was expected that the olefin loss path would permit ethylene

incorporation via trapping of (silox)₃Ta (2) or (silox)₂HTa(κ²-

O,C-OSi^tBu₂CMe₂CH₂) (8), but on the off-chance that the

cyclopentene-derived species were thermodynamically more

stable than those produced from ethylene, the complementary

experiment was conducted. When the ethylene complex was

1-C₂D₄

(silox)₂[^tBu₂(Me₂CCH_3-x/D_x)SiO]NbdC(H/D_y)CH_3-z/D_z

1dCHMe-d₄

(9)

After 12 h, deuterium was indeed distributed in all three sites,

and after 24 h, 96% of the deuterium was present in the silox

c

heated at 103 °C with 56 equiv of C₅H₈present, the tuck-in

ethyl, (silox)₂EtTa(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-Et), and alkyl-

idene 2dCHMe formed with no evidence of cyclopentene-

derived products (eq 13).

2

ligands according to H NMR spectroscopy.

Next, the ¹³C-labeled ethylene complex, (silox)₃Nb(η-¹³C₂H₄)

(1-¹³C₂H₄), was thermolyzed at 155 °C in the presence of ∼5

equiv of nonlabeled ethylene (eq 10).

(silox)₃Ta(η-C₂H₄) + ^cC₅H

₈h 6-Et + ^cC H h

5

8

2-C₂H₄

benzene-d₆

(silox)₃Nb(η-¹³C₂H₄) + C₂H

(silox)₃TadCHMe + ^cC₅H₈

2dCHMe

(13)

₄y

z

155 °C

1-¹³C₂H₄

(silox)₃Nb(η-C₂H₄) + ¹³C₂H₄

1-C₂H₄

(10)

The lack of olefin exchange in the tantalum system rules out

the possibility of olefin loss as a viable mechanistic path. By

inference, and with careful observation of the qualitative

rearrangement and exchange rates in eqs 10 and 11, the olefin

loss mechanism is also eliminated for niobium.

The tuck-in hydride, (silox)₂HTa(κ²-O,C-OSi^tBu₂CMe₂CH₂)

(8) was heated at 103 °C with 10 equiv of ethylene, and after

23 min 57% of the tuck-in ethyl 6-Et was generated along with

2-C₂H₄(26%) and 2dCHMe (3%). Since it is unlikely that 8

reverts to (silox)₃Ta (2) under these conditions,³⁸it is assumed

that the 2-C₂H₄:2dCHMe product ratio is generated from 6-Et.

The 1.6(3) kcal/mol difference given by the 8.7:1 ratio at 103

°C is between the predicted ∆∆G^q≈ 2.7 kcal/mol that

corresponds to a kinetic product ratio of ∼37:1 and the ∆G° of

1.1 kcal/mol, which represents the thermodynamic 4.4:1 ratio.

Once the equilibrium K_alk/oleof ∼0.35 was reached, only 30%

of the ethylene and alkylidene complexes retained the label,

indicating that olefin exchange occurred on a similar time scale

as the rearrangement.⁴⁵When the propene derivative (silox)₃Nb-

(η-C₂H₃Me) (1-C₂H₃Me) was heated in the presence of ∼10

equiv of ethylene (eq 11), about 50% of 1-C₂H₄formed within

27 min and within 64 min, 90% conversion to the ethylene

complex was noted along with 8% (silox)₃NbdCHEt (1dCHEt)

and 2% (silox)₃NbdCHMe (1dCHMe).

(59) Casanova, D.; Cirera, J.; Llunell, M.; Alemany, P.; Avnir, D.; Alvarez, S.

J. Am. Chem. Soc. 2004, 126, 1755-1763.

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4817

A R T I C L E S

Scheme 4

Hirsekorn et al.

The direct H-transfer paths proposed in Scheme 3 can be

differentiated from the double abstraction path by labeling the

R-positions of the olefin adduct with deuterium. The H-transfer

mechanisms lead to a R,â-dideuterio alkylidene product, whereas

double abstraction affords an R-deuterated alkylidene that has

a deuterium incorporated into a silox group. Note that this

experiment can go awry if reversible δ-abstraction by the olefin

R-carbon enables scrambling into the silox groups prior to

rearrangement (Scheme 3, scrambling). Since the 1-butene

conversion conveniently stops at the tuck-in butyl (6-ⁿBu) at

103 °C, R,R-dideuterio-1-butene⁶⁰was chosen as a convenient

assay, as Scheme 4 illustrates. Rearrangement of (silox)₃Ta(η-

D₂CCHEt) (2-D₂CCHEt) at 103 °C afforded the tuck-in R,R-

dideuterio-butyl, (silox)₂(ⁿPrCD₂)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂)

essentially irreversible (the ∼0.0723 k_-2/z_-2in Scheme 4 reflects

a statistical estimate of the sum of reversible CDH₂and CD₂H

δ-abstractions by the alkylidene). By knowing k₂and k_-2from

monitoring, in tandem, the perprotio sample, the isotope effects

may be broken down with simulation of all relevant processes.

The first R-abstraction rate is influenced by a primary KIE (z₂)

and an R-secondary KIE (z₂′) such that z₂z₂′ ) 6.2, but the

second R-abstraction KIEs loss of D from 6-CHDⁿPr-d₂sis

uncomplicated by secondary effects. By ¹H NMR spectroscopic

monitoring of the growth of the alkylidene hydrogen of 2d

CHⁿPr-d₂, a z₂of 4.4 was revealed, a value comparable to other

abstractions,^63-76especially when taking account of the high

temperature. The R-secondary KIE is substantial (z₂′ ) 1.4 )

6.5/4.4) but certainly within reason.

n

(6-CD₂Pr), with no detectable deuterium loss from the R-posi-

The transparency of the KIEs in the tantalum system provides

a benchmark for comparison with the niobium system in which

the lack of an observable tuck-in alkyl intermediate renders

identification of the rate-determining step problematic. The KIE

tions according to ¹H NMR spectroscopy. A significant normal

secondary isotope effect of z₁′ ) 1.23 (z₁′ ) k_H/k_D,^61,62observed

z₁′²) 1.52; the two R-positions were assumed to have the same

secondary effect) was determined for the δ-abstraction event

by comparison of an all-protio sample run in conjunction with

the dideuterio probe.

(63) Fryzuk, M.; Duval, P.; Mao, S.; Zaworotko, M.; MacGillivray, L. J. Am.

Chem. Soc. 1999, 121, 2478-2487; z ) 3.0 at 343 K.

(64) Wood, C. D.; McLain, S. J.; Schrock, R. R. J. Am. Chem. Soc. 1979, 101,

3210-3222; z ) 6.0 at 309 K.

Monitoring the ensuing formation of (silox)₂(silox-d₁)-

TadCDⁿPr (2dCDⁿPr-d₂) at 166 °C was more complicated,

but beneficially so. While the R-abstraction can be treated as

irreversible because the deuterium is lost to 81 equivalent

positions of the three silox groups, the pseudo-reversible process

of δ-abstraction by the alkylidene affords (silox)_2-x(silox-d₁)_x-

(ⁿPrCHD)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂)_1-y(κ²-O,C-OSi^tBu₂-

CMe₂CH₂-d₁)_y(x + y ) 1, 6-CHDⁿPr-d₂), which contains one

(65) Schrock, R. R.; Fellmann, J. D. J. Am. Chem. Soc. 1978, 100, 3359-3370;

z ) 2.7 at 348 K.

(66) Cheon, J.; Rogers, D.; Girolami, G. J. Am. Chem. Soc. 1997, 119, 6804-

6813; z ) 5.2 at 378 K.

(67) McDade, C.; Green, J. C.; Bercaw, J. E. Organometallics 1982, 1, 1629-

1634; z ) 2.9 at 371 K.

(68) Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, 223-233; z )

2.4, 364 K.

(69) Buchwald, S. L.; Nielsen, R. B. J. Am. Chem. Soc. 1988, 110, 3171-

3175; z ) 5.2 at 353 K.

(70) Schock, L. E.; Brock, C. P.; Marks, T. J. Organometallics 1987, 6, 232-

241; z ) 6.5 at 343 K.

n

deuterium in the R-carbon of the Bu and another among the

(71) Luinstra, G. A.; Teuben, J. H. Organometallics 1992, 11, 1793-1801; z

) 5.1 at 353 K, 5.7 at 315 K.

tert-butyl groups of the silox ligands and the cylometalated silox.

(72) Bruno, J. W.; Smith, G. M.; Marks, T. J.; Fair, C. K.; Schultz, A. J.;

Williams, J. M. J. Am. Chem. Soc. 1986, 108, 40-56.

(73) Mayer, J. M.; Curtis, C. J.; Bercaw, J. E. J. Am. Chem. Soc. 1983, 105,

2651-2660; z ) 9.7 at 298 K.

R-Abstraction of deuterium leads to (silox)_3-(2-x)(silox-d₁)_2-2x

-

(silox-d₂)_xTadCHⁿPr (0 < x <1, 2dCHⁿPr-d₂), which is also

(74) Bulls, A. R.; Schaefer, W. P.; Serfas, M.; Bercaw, J. E. Organometallics

1987, 6, 1219-1226; z ) 2.5 at 413 K.

(60) Budzikiewicz, H.; Bold, P. Org. Mass. Spectrom. 1991, 26, 709-712.

(61) Carpenter, B. K. Determination of Reaction Mechanisms; Wiley-Inter-

science: New York, 1984.

(62) Melander, L.; Saunders, W. H., Jr. Reaction Rates of Isotopic Molecules;

Wiley-Interscience: New York, 1980.

(75) Bennett, J. L.; Wolczanski, P. T. J. Am. Chem. Soc. 1997, 119, 10696-

10719; z ) 7.4, 13.7 at 298 K, 5.6 at 363 K.

(76) Schaller, C. P.; Cummins, C. C.; Wolczanski, P. T. J. Am. Chem. Soc.

1996, 118, 591-611; z ) 6.3, 7.1, 4.6 at 370 K.

9

4818 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

A R T I C L E S

Scheme 5

for k_f, which corresponds to the conversion of (silox)₃Nb-

(η-D₂CCHEt) (1-D₂CCHEt) to (silox)₂(silox-d₁)NbdCDⁿPr

(1dCDⁿPr-d₂), was observed to be 2.5 at 166 °C. If forma-

tion of the presumed tuck-in butyl (silox)₂(ⁿPrCD₂)Nb(κ²-O,C-

to the tuck-in secondary alkyl, (silox)_2-x(silox-d₁)_x(PhCH₂CHD)-

((CD₂H)PhHC)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂)_1-y(κ²-O,C-OSi^t-

Bu₂CMe₂CH₂-d₁)_y(x + y ) 1, 6-CHPh(CD₂H)), can cause loss

of label to the silox groups prior to formation of the alkylidene,

as indicated by the scrambling path in Scheme 3. Unfortunately,

since there were no practical independent means of monitoring

the scrambling process, inclusion of these steps into the

simulations do not solve the analytical problems. Nevertheless,

the inability to fit the styrene labeling experiments is strong

evidence for the presence of the scrambling path in this instance.

The tuck-in secondary alkyl 6-CHPh(CD₂H) is unique because

of the R-phenyl substituent that will inductively stabilize the

OSi^tBu₂CMe₂CH₂) (4-CD₂Pr) were rate-determining, the KIE

n

should be e1.5, which was the value in the tantalum system

(i.e., z₁′²) (1.23)²) at 103 °C. For the value of 2.5 to be

attributable to formation of the tuck-in butyl, a secondary isotope

effect of about 1.6 would need to be invoked, and this is outside

the normal range,^61,62especially when taking account of the high

temperature. The value establishes R-abstraction as the rate-

determining step in the Nb double abstraction mechanism. Since

secondary equilibrium isotope effects (z₁′/z_-1′) are expected to

be ∼1.0, the z_fof 2.5 for Nb is approximately z₂z₂′, which in

the tantalum case was 6.2. Conventionally, the lower value for

Nb would be ascribed to a less-symmetric transition state for

H/D transfer. Since the tuck-in butyl is not observed for Nb,

the transition state for conversion to the alkylidene should come

earlier than in the tantalum case where the intermediate and

butylidene are within ∆G° < 1.0 kcal/mol at 166 °C.

The phenyl ring renders the styrene systems different from

the other monosubstituted olefins; thus an attempt to measure

isotope effects was made similar to Scheme 4. A thermolysis

of (silox)₃Ta(η-D₂CCHPh) (2-D₂CCHPh)⁷⁷was conducted in

tandem with 2-C₂H₃Ph. Simulation of the concentration vs time

profiles of the various labeled intermediates failed to yield

tenable rate constants and isotope effects. In reviewing the data,

the more rapid than expected growths of (silox)_2-x(silox-d₁)_x

(PhCH₂CHD)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂)_1-y(κ²-O,C-OSi^tBu₂-

CMe₂CH₂-d₁)_y(x + y ) 1, 6-CHDCH₂Ph-d₂) and (silox)₂(silox-

d₁)Ta(η-DHCCHPh) (2-DHCCHPh-d₂) were responsible for the

difficulty in data fitting. Note that reversible cyclometalation

tantalum alkyl bond, which is polarized Ta^δ+

-

C^δ-HPh(CD₂H). Similar reasoning has been used to explain the

unusual stability of metal-benzyl bonds in early metal C-H

activation systems.^75,76

The absence of a tuck-in alkyl intermediate in the case of

norbornene might be ascribed to a different mechanism.

However, at 103 °C, rearrangement of (silox)₃Ta(η-^cC₇H₈D₂)

(2-^cC₇H₈D₂), which was labeled with deuterium in the olefinic

positions,^78-80afforded the â-deuterio norbornylidene, (silox)₂-

(silox-d₁)Tad^cC₇H₉D (2d^cC₇H₉D-d₂), consistent with the es-

tablished pathway. Had a hydride transfer path been operable,

the â,â-dideuterio norbornylidene would have been produced,

but ¹H and ²H NMR spectroscopy established the â-CH_exoD_endo

group and revealed a deuterium in a silox ligand. The rear-

rangement of 2-^cC₇H₈D₂is complicated but informative, as

Scheme 5 shows. The isotope effect for k_fwas 2.96(3), which

indicated that the R-abstraction step is rate-determining. If

formation of the tuck-in alkyl is construed as two sp²f sp³

(78) Stille, J. K.; Sonnenberg, F. M. J. Am. Chem. Soc. 1966, 88, 4915-4921.

(79) Chamberlin, A. R.; Stemke, J. E.; Bond, F. T. J. Org. Chem. 1978, 43,

147-155.

(77) Fischetti, W.; Heck, R. F. J. Organomet. Chem. 1985, 293, 391-405.

(80) Bond, F. T.; DiPietro, R. A. J. Org. Chem. 1981, 46, 1315-1318.

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4819

A R T I C L E S

Hirsekorn et al.

Scheme 6

changes (π-back-bonding presumably renders the norbornene

olefin carbons between sp²- and sp³-hybridization), each con-

tributes a small inverse isotope effect for the k₁step (δ-

abstraction), but the reverse (k_-1) process should contain two

normal secondary isotope effects.^61,62The EIE for tuck-in alkyl

formation should be inverse. Consequently, the KIE for

R-abstraction, which contains primary and â-secondary isotope

effects (i.e., z₂z₂′′), should actually be >2.96(3), and this value

is clearly in line with that obtained for 1-butene.

Note that the remainder of the scheme for label loss reveals

a stereochemical question. One can assume the loss of label

from the alkylidene of 2d^cC₇H₉D-d₂occurs via a microscopi-

cally reversible path: (1) a CH bond (neglecting R-secondary

effects on 2/81 silox methyl hydrogens, the rate is ∼80k_-2

81z_-2′′) of a silox adds to the endo-face of the alkylidene

(another δ-abstraction, but by the alkylidene) to give (silox)_2-x

/

-

(silox-d₁)_x(^cC₇H₉D_â)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂)_1-y(κ²-O,C-

OSi^tBu₂CMe₂CH₂-d₁)_y(x + y ) 1, 6-^cC₇H₉D_â-d₂); (2) â-ab-

straction occurs only from the exo-face to give (silox)₂(silox-

d₁)Ta(η-^cC₇H₉D) (2-^cC₇H₉D-d₂) with k_-1/z_-1′ as the rate; (3) a

δ-abstraction generates the tuck-in alkyl across the exo-face with

a rate reflecting the 50% probability of correct olefin orientation

(k₁/2z₁); and (4) an essentially irreversible R-abstraction (k₂/z₂)

to afford 2d^cC₇H₁₀-d₂. A fit of the data assuming this path

affords an overall KIE of 1.08(1) consistent with a rate-

determining k_-2step as expected. Addition of nucleophiles to

2-norbornanone occurs toward both faces, with a slight prefer-

ence toward the exo-side;⁸¹hence addition to the exo-face is

possible. However, such a path ultimately leads to an endo-

olefin species, which is expected to be heavily disfavored

relative to the exo-isomer.⁸²

As in the butene rearrangements, the tantalum norbornene

case is helpful in assessing the corresponding niobium chemistry.

Thermolysis of (silox)₃Nb(η-^cC₇H₈D₂) (1-^cC₇H₈D₂) provided the

â-deuterio norbornylidene, (silox)₂(silox-d₁)Nbd^cC₇H₉D (2d

^cC₇H₉D-d₂), consistent with the double abstraction pathway. At

166 °C, the k_H/k_Dof the rearrangement was 2.3(2), a value

commensurate with the butene case, and one best accommodated

by a rate-determining R-abstraction, as explained for the

tantalum rearrangement. If this is true, scrambling of the

â-deuterium of the norbornylidene in 2d^cC₇H₉D-d₂to a silox

group should occur with k_-2(δ-abstraction by the nor-

bornylidene) as the rate-determining step. By fitting the

1-^cC₇H₈D₂to 2d^cC₇H₉D-d₂to d₂-(silox)₃Nbd^cC₇H₁₀(1d

^cC₇H₁₀-d₂) rearrangements and assuming the standard mecha-

nism shown in Scheme 5, the loss of the â-D occurred with a

k_rof 6.17(7) × 10^-6s^-1at 166 °C (∆G_r^q) 36.5(1) kcal/mol).

The rearrangement occurs with a standard free energy change

of -4.5(1) kcal/mol at 166 °C, which is approximately -3.9

kcal/mol at 103 °C, given the usual assumptions.

scrambled 1,6-D₂-cyclohexene complex, (silox)₃Nb(η-^cC₆H₈-

1,6-D₂) (1-^cC₆H₈-1,6-D₂), which was formed upon â-H elimina-

tion, to the alkylidene is roughly 20:1. That product ratio is

approximately the kinetic ratio k_-1/k₂/z₂, and if the primary

isotope for abstraction is the same as that in the norbornene

case (z₂≈ 2.4 at 103 °C), then the product ratio is ∼8.3 and

q

∆∆G^q) ∆G₂- ∆G_-1≈ 1.6 kcal/mol. Unfortunately,

analytical difficulties prevented the extraction of isotope effects

in the Nb case, but the qualitative message is clear.

In the case of tantalum, Table 2 indicates that the rearrange-

ment of (silox)₃Ta(η-^cC₆H₈-1,2-D₂) (2-^cC₆H₈-1,2-D₂) to (silox)₂-

(1,2-D₂-^cC₆H₈)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-^cC₆H₈-1,2-D₂)

is essentially irreversible; thus no deuterium is expected to

scramble in the cyclohexene complex since rearrangement of

6-^cC₆H₈-1,2-D₂to (silox)₂(silox-d)Tad^cC₆H₉-2-D-d₂(2-^cC₆H₉-

2-D-d₂) occurs more quickly. While 6-^cC₆H₈-1,2-D₂could not

2

be assayed with confidence, H NMR spectroscopy confirmed

that the deuterium labels were in the silox ligand and the

â-position of the 2-^cC₆H₉-2-D-d₂, and no deuterium scrambling

in the cyclohexene complex was detected.

Equilibria in the Niobium System. 1. Olefin Exchange. In

contrast to tantalum, niobium olefin complexes (silox)₃Nb(ole)

(1-ole) typically undergo substitution with free olefins faster

than their rearrangements, or at least on the same time scale.⁴⁵

By measuring various equilibria, the relative stabilities of the

olefin and alkylidene complexes were established and compared

with those calculated for the models (HO)₃Nb(ole) (1′-ole).

Where possible, 1-ole complexes were subjected to concentra-

tions of free olefin that enabled measurement of K_ole′/ole

)

[1-ole′][ole]/[1-ole][ole′] by direct integration of all four species

If the R-abstraction is the rate-determining step in the

rearrangement of (silox)₃Nb(η-^cC₆H₈-1,2-D₂) (1-^cC₆H₈-1,2-D₂),

the deuterium labels in the olefinic positions⁸³should scramble

within the cyclohexene prior to or at least competitively with

formation of the alkylidene, (silox)₂(silox-d)Nbd^cC₆H₉-2-D-

d₂(1d^cC₆H₉-2-D-d₂), as Scheme 6 illustrates. At early conver-

sion, where ∼1% of the alkylidene has formed, the ratio of the

(eq 14).

103 °C

(silox) Nb(ole′) + ole

y

z

(14)

3

(silox)₃Nb(ole) + ole′

1-ole

1-ole′

With little competing formation of the respective alkylidene

compounds, a number of equilibria could be measured in this

fashion, as Table 3 shows. Only the relative energy of

(silox)₃Nb(η-^cC₆H₁₀) (1-^cC₆H₁₀) could not be determined, but

given the accuracy of the relatiVe energies of the computational

(81) Ohwada, T. Tetrahedron 1993, 49, 7649-7656.

(82) (a) Clavell, K. J. Coord. Chem. ReV. 1996, 155, 209-243. (b) Yamamoto,

Y.; Ohno, T.; Itoh, K. Organometallics 2003, 22, 2267-2272.

(83) Ahlgren, G.; Akermark, B. Acta Chem. Scand. 1968, 22, 1129-1132.

9

4820 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

A R T I C L E S

Table 3. Equilibrium Measurements (K, ∆G° (kcal/mol) for (silox)₃Nb(ole) (1-ole) + ole′ h (silox)₃Nb(ole′) (1-ole′) + ole (103 °C, ∆G°(ole′/

ole), eq 14) and (silox)₃Nb(ole) (1-ole) + ole′ h (silox)₃Nb(alk′) (1dalk′) + ole (155 °C, ∆G°(alk′/ole), eq 15, Corrected to 103 °C);

Calculated Equilibria at 103 °C

∆

G

°

∆

(ole

G_calcd

°

∆

(alk

G₁₅₅

°

∆

(alk

G₁₀₃

°

∆

(alk

G_calcd

°

reaction

K

(ole

′

/ole)

′

/ole)^a

′

/ole)

′

/ole)

′

/ole)^b

1-C₂H₃Et + C₂H₃Me h 1-C₂H₃Me + C₂H₃Et

1-C₂H₃Et + C₂H₃Ph h 1-C₂H₃Ph + C₂H₃Et

1-C₂H₃Et + ^cC₇H₁₀h 1-^cC₇H₁₀+ C₂H₃Et

1-^cC₅H₈+ C₂H₃Et h 1-C₂H₃Et + ^cC₅H₈

1-C₂H₄+ C₂H₃Ph h 1-C₂H₃Ph + C₂H₄

0.76(1)

17.3(5)

2.18(5)

24.0(2)

0.0367(30)

9.73(8)

0.2

0.0

-2.1

-0.58

-2.37

2.5

-1.7

0.55

-2.0

-0.1

-3.0

2.5

c

t

1-C₂H₃Bu + C₂H₃Et h 1-C₂H₃Et + C₂H₃Bu

1-C₂H₃Ph + C₂H₃Ph-p-OMe h 1-C₂H₃Ph-p-OMe + C₂H₃Ph

1-C₂H₃Ph + C₂H₃Ph-p-CF₃h 1-C₂H₃Ph-p-CF₃+ C₂H₃Ph

1-C₂H₄+ C₂H₃Me h 1dCHEt + C₂H₄

0.476(2)

5.11(4)

d

e

-1.22

0.209(4)

0.390(3)

0.157(6)

0.005(1)

0.162(9)

1.33

0.80

1.57

4.5

1.85

1.32

2.09

5.0

8.9

8.8

9.7

9.6

c

1-C₂H₄+ C₂H₃Et h 1dCHⁿPr + C₂H₄

1-C₂H₄+ ^cC₅H₈h 1d^cC₅H₈+ C₂H₄

1-C₂H₄+ ^cC₆H₁₀h 1d^cC₆H₁₀+ C₂H₄

t

1-C₂H₄+ C₂H₃Bu h 1dCHCH₂Bu + C₂H₄

1.55

2.07

^aCalculations have an average deviation of -0.1(4) kcal/mol. ^bCalculations have an average deviation of +6.7(14) kcal/mol from the observed. ^cCalculations

t

were not performed on (HO)₃Nb(C₂H₃Bu) (1′-C₂H₃Bu or 1′dCHCH₂Bu). ^dCalculations were not performed on (HO)₃Nb(C₂H₃Ph-p-OMe) (1′-C₂H₃Ph-

p-OMe). ^eCalculations were not performed on (HO)₃Nb(C₂H₃Ph-p-CF₃) (1′-C₂H₃Ph-p-CF₃).

Table 4. Relative Observed, Estimated, and Calculated ((HO)₃Nb(olef) (1′-ole); (HO)₃Nb(alk) (1′)alk)) Standard Free Energies in kcal/mol

((0.2) at 103 °C for (silox)₃Nb(ole) (1-ole), (silox)₃Nb ) (alk) (1dalk), and the (silox)₃Nb(ole) (1-ole) h (silox)₃Nb(alk) (1dalk)

Rearrangement (Scheme 1); Activation Free Energies for ∆G^q(k_f) and ∆G^q(k_r) Are Included for Determination of Figure 4

reaction

∆G

°

(ole)^a

∆G

°

(alk)^b

∆

G^q(k_f)

∆

G^q(k_r)

∆G

°

(alk/ole)

∆G_calcd

°

(ole)^c

∆G_calcd° (alk/ole)

1-C₂H₄h 1dCHMe

0.0

1.6

3.2

0.1

1.3

1.9

2.1

5.0

35.5

31.0

32.7

33.4

32.7

31.5

29.5

33.9

34.8

34.9

36.0

36.3

36.9

36.6

1.6

0.7

0.0

2.5

4.4

4.5

e

2.6

8.5

-0.3

4.3

e

2.1

-1.4

1-C₂H₃Ph h 1dCHCH₂Ph

1-^cC₇H₁₀h 1d^cC₇H₁₀

1-C₂H₃Et h 1dCHⁿPr

1-C₂H₃Me h 1dCHEt

2.5

4.0

4.6

4.8

6.3

7.2

-3.9

-3.3

-2.9

-4.2

-5.1

t

1-C₂H₃Bu h 1dCHCH₂Bu^d

1-^cC₅H₈h 1d^cC₅H₈

7.5

10.9

1-^cC₆H₁₀h 1d^cC₆H₁₀

>8.4^f

>32.9^f

>-3.4^f

^aStandard free energy of 1-ole relative to that of 1-C₂H₄at 0.0 kcal/mol. ^bStandard free energy of 1dalk relative to that of 1-C₂H₄at 0.0 kcal/mol.

^cCalculated standard free energy of 1′-ole relative to that of 1′-C₂H₄at 0.0 kcal/mol; the average deviation is 0.1(3) kcal/mol. ^dEstimates of the standard

t

free energy of (silox)₂(^tBuCH₂CH₂)Nb(κ²-O,C-OSi^tBu₂CMe₂CH₂) (4-CH₂CH₂Bu) place it ∼2.1 kcal/mol. ^eCalculations were not performed on

t

(HO)₃Nb(C₂H₃Bu) (1′-C₂H₃Bu or 1′dCHCH₂Bu). ^fAssuming the minimum amount of a substance that can be observed with confidence by ¹H NMR

spectroscopy is 3%.

models pertaining to the remaining olefin complexes, its energy

can be estimated with confidence.

The latter measurement was given more weight; thus a

compromise value of +10 was used to estimate cases needing

entropy corrections.

The relative standard free energy of the norbornylidene,

(silox)₃Nbd^cC₇H₁₀(1d^cC₇H₁₀), could not be determined via

equilibrium measurements because of interference from polym-

erization of the norbornene. However, in this circumstance, the

KIE experiments above enabled the activation free energy for

k_rto be determined (vide supra), and upon correcting to 103

°C, ∆G° ) ∆G_f^q- ∆G_r^q) -3.9 kcal/mol.

2. Alkylidenes Relative to Olefin Complexes. Aside from

the niobium ethylene and styrene rearrangments, the standard

free energy changes could not be determined because the

alkylidene proved to be too exoergic for the equilibria to be

measured by NMR spectroscopic integration. In these instances,

the relative energies of the alkylidenes could be obtained by

measuring equilibria between lower energy olefin complexes

and the 1-alk of interest via eq 15.

The data in Table 3 provide experimental relative standard

free energies of (silox)₃Nb(ole) (1-ole) and (silox)₃Nb(alk) (1-

alk) listed in Table 4, with the exceptions noted above. The

smallest olefin, ethylene, and styrene, which possesses the

electron-withdrawing Ph substituent, are the most tightly bound.

Next is norbornene, presumably due to the diminution of ring

strain^84,85as back-bonding renders it more sp³-like. 1-Butene

and propene follow and bind with similar energies, while the

hindered tert-butylethylene is more weakly attached. The

hindered cyclic cases are the most weakly bound, with the

smaller cyclopentene bound more tightly than cyclohexene by

at least 1.2 kcal/mol, and more likely by around 3.7 kcal/mol,

since the calculated number in this case is probably accurate.

This is a typical list for the electron-rich Nb(III), with sterics

favoring the smaller olefins, and olefins with electron-withdraw-

155 °C

(silox) Nb(ole) + ole′

(silox) Nb(alk′) + ole

y

z

(15)

3

1-ole

1-alk

At 155 °C, where the rearrangement is swifter, a judicious

selection of free olefin permitted measurement of K_alk′/ole

)

[1-alk′][ole]/[1-ole][ole′] by direct integration of all components,

as Table 3 reveals. To correct to 103 °C, the standard entropy

changes must be known. Epoxide to aldehyde/ketone isomer-

izations were considered as structural models that typically

possess ∆S° values that range from +4 to +10 eu,⁸⁴and the

measured ∆S° for the niobium ethylene case was +12.5 eu.

(84) For many rudimentary thermodynamics estimates, consult: http://

webbook.nist.gov/chemistry/. For example, using Benson additivities (ref

85) and appropriate data, we found that the strain energies (E_s((1-2 kcal/

mol)) of pertinent molecules are: ^cPeH, 6.5; C₅H₈, 6.0; methylenecyclo-

c

pentane, 5.3; ^cHexH, 0.3; ^cC₆H₁₀, 1.6; cyclohexanone, 1.0; norbornane, 15.6;

norbornene, 23(7); and 2-norbornanone, 14.3.

(85) Benson, S. W. Thermochemical Kinetics; Wiley & Sons: New York, 1968.

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4821

A R T I C L E S

Hirsekorn et al.

ing substituents being favored over those with greater electron

donating capacity.

the time scale of rearrangement, and a series of labeling

experiments ruled out the simplest path, that of H-transfer. A

competing reaction in which δ-abstraction leads to a tuck-in

secondary alkyl can also be eliminated on the basis of labeling

experiments for most substrates. However, the inability to fit

rate data for the rearrangement of (silox)₃Ta(η-D₂CCHPh) (2-

D₂CCHPh) is best accommodated in view of a competing

scrambling path involving a tuck-in secondary alkyl intermedi-

ate. Rearrangements of the cyclic olefinsscyclopentane, cy-

clohexane, and norbornenesrequire the intermediacy of a tuck-

in secondary alkyl species. It is surprising that secondary

alkylidenes derived from acyclic substrates were not identified,

especially since the energies of the cyclic alkylidenes suggest

that such species should be close to (silox)₃MdCHR. It is

doubtful that the transition state leading from a putative tuck-

in secondary alkyl to secondary alkylidene would be problematic

because rearrangements of the cyclics are relatively swift. It is

more likely that the secondary akylidenes are unstable with

respect to their primary counterparts. In the tantalum system,

the tuck-in primary alkyls are the energetically dominant species;

thus any tuck-in secondary alkyl may simply be relatively too

high in energy to be easily discerned. For niobium, evidence

for a tuck-in secondary alkyl was indirect; the 2-butene complex

isomerized to the 1-butene species prior to the latter’s rear-

rangment to butenylidene. Extended thermolyses (170 °C, >10

d) of (silox)₃Nb(ole) (ole ) propene, 1-C₂H₃Me; 1-butene,

1-C₂H₃Et; styrene, 1-C₂H₃Ph) failed to elicit any evidence of

secondary alkylidenes. Since their formation is kinetically viable,

it can be concluded that such species are roughly 4 kcal/mol

above primary alkylidenes in energy.

3. Calculations. Table 3 indicates that the equilibria of eq

14 were calculated fairly well, with an average deviation of

-0.1(4) kcal/mol, but the experimental vs calculational olefins

to alkylidene equilibria of eq 14 are poorly matched. If the

calculated ∆G°((HO)₃Nb(ole), 1′-ole) values are referenced to

1-C₂H₄at 0.0 kcal/mol (Table 4), they match the experimental

relative olefin energies quite well and possess an average

deviation of 0.1(3) kcal/mol. This is expected, since it is

basically complementary to the equilibria of eq 13. Likewise,

the olefin/alkylidene equilibria for each substrate are poorly

calculated, although a correction factor of -6.2(2.9) (-7.5(3)

with the ethylene case left out) to each ∆G_calcd°(alk/ole) would

provide a decent fit to the experimental data. One possible

explanation is that the olefin binding free energies, although

relatively precise, may be overestimated: the calculated ∆G°-

(1′-ole) values (kcal/mol, relative to (HO)₃Nb (1′) + ole) are

1′-C₂H₄, -27.7; 1′-C₂H₃Ph, -25.2; 1′-^cC₇H₁₀(1′-nor), -23.3;

1′-C₃H₆, 1′-C₄H₈, -23.2; 1′-^cC₅H₈, -20.2; 1′-^cC₆H₁₀, -16.8.

Preliminary olefin substitution kinetics suggests that the free

energy of ethylene binding in 1-C₂H₄must be >-24 kcal/mol,⁴⁵

which suggests that the calculated ethylene binding energy in

1′-C₂H₄is too favorable. However, in the three measured cases

(1-C₂H₄f 1dCHMe, ∆H° ) 6.3; 2-C₂H₄f 2dCHMe, ∆H°

) 7.8; 2-C₂H₃Ph f 2dCHCH₂Ph, ∆H° ) 10.4 kcal/mol), the

calculated standard enthalpy change underestimated these values

by -0.6, -3.4, and -2.6 kcal/mol. This is surprising, because

steric factors are likely to attenuate the binding of olefins to a

greater extent than the alkylidenes, and these influences are

obviously not accounted for in the calculational model.⁸⁶

While the olefin to alkylidene entropy change would be

expected to favor the alkylidene, and this is corroborated in the

measured cases (1-C₂H₄f 1dCHMe, ∆S° ) 12.5(10) eu;

2-C₂H₄f 2dCHMe, ∆S° ) 16.6(16) eu; 2-C₂H₃Ph f

2dCHCH₂Ph, ∆S° ) 25(1) eu), additional degrees of freedom

in the alkylidene fragment vs the olefin do not contribute enough

to account for the standard entropy change observed. The

calculated entropy changes for the same cases are + 8.2, +5.3,

and +2.1 eu, respectively. From these discrepancies, it is likely

that entropic contributions from the silox groups upon rear-

ranging from olefin to alkylidene are a principal reason the

calculations are in error.

87

Calculations performed on :CHEt and :CMe₂provide a

plausible reason for the relative instability of the secondary

alkylidenes. The singlet ground state of :CMe₂is ∼8.6 kcal/

mol below that of the primary carbene, whose ground state is

actually a triplet by ∼1.1 kcal/mol. Hyperconjugation of the

six CH bonds into the empty p-orbital of the singlet state renders

it substantially lower in energy than :CHEt. These same

interactions may tend to destabilize any secondary metal-

alkylidene because the p-orbital is now involved in π-bonding

and 4e^-interactions from the CH bonds will now be repulsive

in character. Entropic factors pertaining to the silox groups may

also disfavor secondary alkylidenes.

Scheme 7 illustrates the transition state for δ-abstrac-

tion^68,72,88-90leading to the tuck-in alkyl intermediate and the

subsequent R-abstraction^22,23,63-69transition state that leads to

the alkylidene product. The δ-abstraction (k₁) transformation

is unusual and deserves additional comment, because one could

invoke oxidative addition of a silox Me-group to the nominal

Nb(III) center followed by olefin insertion in lieu of this single

step. First, although structural details are only known for

(silox)₃Ta(η-CH₂CHEt) (2-C₂H₃Et), it is likely that related

niobium olefin complexes have a substantial amount of meta-

Discussion

Olefin to Alkylidene Rearrangement. 1. Mechanism. The

mechanistic possibilities outlined in Scheme 3 have been pared

by experiment, leaving the double abstraction pathway as the

means by which (silox)₃M(ole) (M ) Nb, 1-ole; Ta, 2-ole)

rearranges to (silox)₃M(alk) (M ) Nb, 1dalk; Ta, 2dalk). In

coming to this conclusion, the niobium system is assumed to

behave in a manner similar to tantalum, but with energetic

differences. The most evident difference is the stability of the

tuck-in alkyl intermediate, (silox)₂RM(κ²-O,C-OSi^tBu₂CMe₂-

CH₂) (M ) Nb, 4-R; Ta, 6-R), which is observed only for ole

) tert-butylethylene in the niobium system, but is present in

all but the norbornene rearrangement for tantalum.

(87) Calculations at the G3MP2B3 level at 298.15 K give the following:

∆H(S-T, :CMe₂) ) -2.7 kcal/mol, ∆H(S-T, :CHEt) ) 1.1 kcal/mol,

∆H(S(:CMe₂) - S(:CHEt)) ) 8.6 kcal/mol, ∆H(T(:CMe₂) - T(:CHEt))

) 4.8 kcal/mol. Wilcox, C. F., personal communication.

(88) For other unusual abstractions, see: (a) Wada, K.; Pamplin, C. B.; Legzdins,

P.; Patrick, B. O.; Tsyba, I.; Bau, R. J. Am. Chem. Soc. 2003, 125, 7035-

7048. (b) Ng, S. H. K.; Adams, C. S.; Hayton, T. W.; Legzdins, P.; Patrick,

B. O. J. Am. Chem. Soc. 2003, 125, 15210-15223.

The olefin loss pathway was eliminated by showing that

excess olefin did not exchange into the tantalum system during

(89) Chamberlain, L. R.; Kerschner, J. L.; Rothwell, A. P.; Rothwell, I. P.;

Huffman, J. C. J. Am. Chem. Soc. 1987, 109, 6471-6478.

(90) Fryzuk, M. D.; Johnson, S. A.; Rettig, S. J. J. Am. Chem. Soc. 2001, 123,

1602-1612.

(86) Cundari, T. R.; Gordon, M. S. J. Am. Chem. Soc. 1991, 113, 5231-5234.

9

4822 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

A R T I C L E S

Scheme 7

lacyclopropane character;^46,47that is, that of Nb(V), given the

highly reducing nature of the early metal center. Second, it is

pertinent to view the overall process from the standpoint of the

tuck-in alkyls, (silox)₂(R)M(κ²-O,C-OSi^tBu₂CMe₂CH₂) (M )

Nb, 4-R; M ) Ta, 6-R), partitioning between olefin complex

(k_-1) and alkylidene formation (k₂). The former process is a

â-abstraction (k_-1) reaction common to early metal systems,

especially those that are hindered, as Buchwald^69,91and Bon-

cella,⁹²among others,^70-73,90have intimated. The δ-abstraction

is simply the microscopic reverse of this well-established

process.

2. Niobium Energetics. Without direct evidence of the

intermediate (silox)₂(R)M(κ²-O,C-OSi^tBu₂CMe₂CH₂) (M ) Nb,

4-R), the niobium energetics are limited to assessments of olefin

and alkylidene complex ground states and the rate-determining

transition state. Figure 4. illustrates the energies of these

respective states for each case relative to the energy of

(silox)₃Nb(η-C₂H₄) (1-C₂H₄), which is assigned a value of 0.0

kcal/mol. Generally, the rate-determining transition state energies

increase as the energies of the ground-state olefin complexes

increase, but to a lesser extent. In contrast, the product

alkylidenes are relatively similar in energy, with the exceptions

of (silox)₃Nbd^cC₇H₁₀(1d^cC₇H₁₀) and (silox)₃Nbd^cC₆H₁₀(1d

^cC₆H₁₀). The ground state of (silox)₃Nb(η-^cC₆H₁₀) (1-^cC₆H₁₀)

is given a value of >8.4 kcal/mol, and this inequality is

designated by an arrow in Figure 4, as is the transition state for

rearrangement because the ∆G_f^qof 29.5 kcal/mol is fixed. The

magnitudes of the niobium rearrangement activation energies

are substantial and indicate that even with an internal “catalyst”,

the cyclometalating silox ligand, the conversion of olefin to

alkylidene in the absence of Brønsted or Lewis acids is

energetically costly.

Note also the similarity of the transition states for δ-abstrac-

tion/â-abstraction and R-abstraction/δ-abstraction (by the alkyl-

67,68,70,93-96

idene, k_-2

)

shown in Scheme 7. R-Abstraction is a

common reaction of sterically encumbered early metal cen-

ters^22,23and is distinguished from R-elimination/reductive

elimination paths because the latter cannot occur at d⁰metal

centers. The KIEs discerned for R-abstraction by the tuck-in

alkyls of the 1-butene and norbornene rearrangements are quite

comparable to previous cases^63-69and support the notion that

the R-abstraction step (k₂) is rate-determining for the niobium

system and for (silox)₃Ta(η-^cC₇H₁₀) (2-^cC₇H₁₀). Given the

similarity in TSs shown in Scheme 7 and the extreme crowding

at the metal center that an alternative, six-coordinate, d⁰(silox)₂-

(olefin)HM(κ²-O,C-OSi^tBu₂CMe₂CH₂) intermediate would suf-

fer, formation of the tuck-in alkyl (M ) Nb, 4-R; M ) Ta,

6-R) is best construed as occurring via a δ-abstraction by the

â-carbon of the olefin on a silox methyl group.

The gross features of Figure 4 suggest a linear free energy

(LFE) relationship for the rearrangement. Given the accuracy

of the calculations in assessing relative ground-state olefin

complex energies, (silox)₃Nb(η-^cC₆H₁₀) (1-^cC₆H₁₀) was assigned

the calculated value of 10.9 kcal/mol. Figure 5 illustrates two

related LFEs: ∆∆G_f^q) R - â∆∆G°(1-ole) and ∆∆G_f^q

)

R′ + â′∆∆G°(1dalk/1-ole). As espoused above, the activa-

tion energies for rearrangement are highly dependent on the

ground-state energy of (silox)₃Nb(ole) (1-ole), with â ) 0.56

(R²) 0.74) for the entire set of substrates, and â ) 0.59 (R²)

0.93) with the outlying norbornene point removed. The nor-

bornene is certainly a unique case. While it is tempting to place

it in a class with the other cyclics, 1-^cC₇H₁₀is 3.2 and ∼6.9

kcal/mol lower than the cyclopentene and cyclohexene com-

plexes, respectively, an observation consistent with its greater

relief of ring strain^84,85upon binding. Note that its ∆G_f^qis quite

similar to the other cyclics, rendering it an outlying point

because the ground state of 1-^cC₇H₁₀is roughly 4 kcal/mol

(91) Buchwald, S. L.; Kreutzer, K.; Fisher, R. A. J. Am. Chem. Soc. 1990, 112,

4600-4601.

(92) (a) Wang, S.-Y. S.; Abbound, K. A.; Boncella, J. M. J. Am. Chem. Soc.

1997, 119, 11990-11991. (b) Wang, S.-Y. S.; VanderLende, D. D.;

Abboud, K. A.; Boncella, J. M. Organometallics 1998, 17, 2628-2835.

(c) Wang, S.-Y. S.; Abboud, K. A.; Boncella, J. M. Polyhedron 2004, 23,

2733-2749.

(93) Chamberlain, L. R.; Rothwell, I. P.; Huffman, J. C. J. Am. Chem. Soc.

1986, 108, 1502-1509.

(94) Coles, M. P.; Gibson, V. C.; Clegg, W.; Elsegood, M. R. J.; Porelli, P. A.

J. Chem. Soc., Chem. Commun. 1996, 1963-1964.

(95) van der Heijden, H.; Hessen, B. J. Chem. Soc., Chem. Commun. 1995,

145-146.

(96) Vaughan, W. M.; Abboud, K. A.; Boncella, J. M. J. Am. Chem. Soc. 1995,

117, 11015-11016.

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4823

A R T I C L E S

Hirsekorn et al.

Figure 4. Relative standard free energies (kcal/mol, 103 °C) of (silox)₃Nb(ole) (1-ole) indicated by bottom of solid blue column and those of (silox)₃-

Nb(alk) (1dalk) indicated by bottom of striped blue column. The solid blue column is the activation free energy for the overall forward step in the rearrangement,

∆G_f^q, while the striped blue column refers to ∆G_r^q. The common top of the columns refers to the transition state energy of the rate-determining step. All

energies are relative to (silox)₃Nb(η-C₂H₄) (1-C₂H₄) at 0.0 kcal/mol. The arrows indicate that the columns represent lower limits for the ground state of

(silox)₃Nb(η-^cC₆H₁₀) (1-^cC₆H₁₀) and its accompanying transition state.

Figure 5. Linear free energy relationships: (1) the relation of the ∆G_f^q’s for the (silox)₃Nb(ole) (1-ole) to (silox)₃Nb(alk) (1dalk) conversion to the relative

standard free energies of 1-ole given as ∆∆G_f^q) R - â∆∆G°(1-ole); (2) the relation of the ∆G_f^q’s to the standard free energy change of the 1-ole to 1dalk

conversion given as ∆∆G_f^q) R′ + â′∆∆G°(1dalk/1-ole). The red line has the (silox)₃Nb(η-^cC₇H₁₀) (1-^cC₇H₁₀) point removed, and the blue line has the

1-^cC₇H₁₀and (silox)₃Nb(η-^cC₆H₁₀) (1-^cC₆H₁₀) points left out. The calculated ∆G°(1-^cC₆H₁₀) of 10.9 kcal/mol was used, but it is likely to approach 13.6

kcal/mol. See text for explanation.

lower. The styrene case, while not dropped in either LFE,

appears to have a slightly higher than expected ∆G_f^qor higher

∆G°(1-C₂H₃Ph). Since it is bulkier than other monosubstituted

olefins, yet binds more strongly, the latter is unlikely. In the

conversion to the transient (silox)₂(PhCH₂CH₂)Nb(κ²-O,C-OSi^t-

Bu₂CMe₂CH₂) (4-CH₂CH₂Ph) complex, it is the â-carbon-

niobium bond that is broken, yet this is the olefin carbon

containing the inductively withdrawing Ph group.^75,76It is likely

that this TS is achieved via an elongated reaction coordinate,

i.e., the d(Nb-C_â(Ph)) is shorter than comparative d(Nb-C_â-

(R)), leading to a higher energy transition state (vide infra). For

this feature to impact Figure 4, the styrene rearrangement may

be the only case in which tuck-in alkyl formation is rate-

determining.

The second LFE relationship correlates ∆G_f^qwith the standard

free energy change of the reaction, ∆G°(1dalk/1-ole), a more

conventional comparison that yields a â of 0.73 (R²) 0.87).

In viewing this correlation, the cyclohexene and norbornene

cases are moderately outlying, and when removed a â of 0.60

(R²) 0.96) is obtained. As stated above, the ∆G°(1-^cC₇H₁₀) is

anomalously low, presumably because of relief of ring strain,

and the norbornylidene 1d^cC₇H₁₀is only modestly lower in

energy than the remaining alkylidenes (∼2 kcal/mol); as such

the ∆G°(1-^cC₇H₁₀/1d^cC₇H₁₀) in this case is higher than

expected. Ring strain is also relieved in formation of the

norbornylidene,^84,85but it is difficult to estimate the extent to

which the sterics of this largest substrate offset this factor. For

cyclohexene, it is unlikely that the slight displacement from the

9

4824 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

A R T I C L E S

cannot be equilibrated, the relative energies of the species must

be estimated, (2) the tuck-in alkyls, (silox)₂RTa(κ²-O,C-

OSi^tBu₂CMe₂CH₂) (6-R), are typically the most stable species

and dominate the energetics, and (3) 6-R and the product

alkylidene (silox)₃Tadalk (2dalk) are often so stable that the

relative energy of corresponding olefin complex (silox)₃Ta(ole)

(2-ole) cannot be discerned.

Since the calculated energies of (HO)₃Nb(ole) (1′-ole) closely

matched the experimental values of (silox)₃Nb(ole) (1-ole), a

reasonable amount of confidence can be placed on the corre-

sponding tantalum olefin complex energies. Figure 7 illustrates

the relative energies of all ground states and transition states in

the tantalum system referenced to (silox)₃Ta(η-C₂H₄) (2-C₂H₄)

at 0.0 kcal/mol; 2-ole is arranged according to the calculated

relative energies of (HO)₃Ta(ole) (2′-ole). Note that the rear-

rangements of the propene, 1-butene, and cyclohexene com-

plexes are undetermined because 2-ole cannot be observed at

equilibrium. In regard to these specific cases, energies pertain-

ing to the 2dalk, 6-R, and their connecting transition states

are indicated as maxima in the graph through labeling with

arrows.

Figure 7 shows that the tuck-in alkyls, (silox)₂RTa(κ²-O,C-

OSi^tBu₂CMe₂CH₂) (6-R), are the most stable species for the

acyclic olefin cases. For norbornene, 6-^cC₇H₁₁is not observed,

and for cyclopentene and cyclohexene, the 6-^cC₅H₉and 6-^cC₆H₁₁

complexes are 1.0 and 0.6 kcal/mol above their respective

alkylidenes 2d^cC₅H₉and 2d^cC₆H₁₁. Secondary alkyl 6-R

species are expected to be higher in energy than primary alkyl

6-R derivatives on the basis of electronic effects. Studies of

(silox)₂(^tBu₃SiNH)Ti-R suggest that a 1.3-2.6 kcal/mol dif-

ference is plausible.⁷⁵In addition, they should have the most

severe steric interactions of any cyclometalated intermediates.

The tuck-in primary alkyls (6-R, R ) Et, CH₂CH₂Ph, ⁿPr, ⁿBu)

are about 2.0-2.5 kcal/mol more stable than their respective

alkylidenes; hence the increase in energy of tuck-in secondary

alkyls is roughly 3 kcal/mol. A similar steric penalty for

secondary alkylidenes might contribute to their higher energy

in relation to their more stable primary counterparts and 6-R

(Ta).

Figure 6. General reaction coordinate vs energy diagram for the (silox)₃-

M(ole) (M ) Nb, 1-ole; Ta, 2-ole) to (silox)₂RM(κ²-O,C-OSi^tBu₂CMe₂-

CH₂) (M ) Nb, 4-R; Ta, 6-R) to (silox)₃M(alk) (M ) Nb, 1dalk; Ta, 2d

alk). The Ta and Nb olefin and alkylidene complexes are portrayed with

the same energy surfaces for conVenience, and are not meant to imply they

exist at the same absolute energies.

remaining cases is due to its ∆G_f^qor ∆G°(1-^cC₆H₁₀), because

the previous correlation was excellent. Cyclohexylidene 1d

^cC₆H₁₀is not as low in energy as the remaining alkylidenes,

and while this could be the origin of the displacement from the

line, it is more likely that the ∆G° for this case is still

undervalued at -5.9 kcal/mol and might be as much as -8.6 if

the calculated values are corrected as rationalized above.

The LFE relationship is consistent with the proposed tuck-in

alkyl to alkylidene transformation as the rate-determining step,

as Figure 6 illustrates. Electronic differences in the olefin should

not have substantial impact on the energies of (silox)₂RNb-

(κ²-O,C-OSi^tBu₂CMe₂CH₂) (4-R) or the alkylidenes (silox)₃-

Nbd(alk) (1dalk); hence the transition state connecting these

two species should also be relatively free of the influence of

olefin substituents. The olefin complexes are sensitive to

electronic changes, and the transition states connecting them to

the tuck-in alkyl, 4-R, should also be influenced by similar

amounts of energy. If this first transition state were rate-

determining, then â would be quite small. Its magnitude is far

more consistent with a rate-determining second transition state.

As the energies of 1-ole change, energies of the transition state

linking 4-R to 1dalk are relatively insensitive, permitting a

substantial fraction of ∆∆G°(1-ole) to be transposed to ∆∆G_f^q.

A destabilization of the olefin complex ground state leads to

swifter rates of rearrangement. Activation energies (∆G_r^q’s) of

the reverse rearrangement should also be less sensitive to

substituent. Table 4 reveals that this is the case, with ∆G_r^q(ave)

) 35.6 (10) kcal/mol representing a modest range of 33.9-

36.9 kcal/mol (34.8-36.9 with the ethylene case excluded)

compared to ∆G_f^q(ave) ) 32.7 (21) with its much greater range

of 29.5-35.5 kcal/mol.

As Figure 7 shows, by dominating the energetics, the tuck-

in alkyls (silox)₂RTa(OSi^tBu₂C-Me₂CH₂) (6-R) change the

importance of the two transition states relative to the niobium

rearrangements. The R-abstraction (k₂) step of the (silox)₃Ta-

(η-C₂H₄) (2-C₂H₄) rearrangement is the slowest, and the second

transition state linking 6-Et and (silox)₃TadCHMe (2dCHMe)

is clearly the highest. In the norbornene case, the KIE for the

overall rearrangement of 2-^cC₇H₁₀to 2d^cC₇H₁₀is best accounted

for in terms of a rate-determining k₂step, just as in all the

niobium cases. However, it appears that in the remaining

tantalum rearrangements, the first transition state linking 2-ole

to 6-R is the highest because of the influence of ∆G°(6-R). For

the sake of argument, assume the ∆G° for 2-ole h 2dalk is

the same as that in the corresponding niobium cases. In support,

the calculated ∆G°(2′-ole/2′-alk) values were effectively the

same for niobium and tantalum. In certain cases where ambigu-

ities exist (propene, 1-butene), 6-ⁿPr, 6-ⁿBu, 2dCHEt, and 2d

CHⁿPr can be assigned relative standard free energies of -0.4,

-0.7, 1.7, and 1.3 kcal/mol in place of the arrows in Figure 7.

As a consequence, the transition states for 6-ⁿPr h 2dCHEt

and 6-ⁿBu h 2dCHⁿPr are at ∆G° ) 32.1 and 31.9 kcal/mol.

3. Tantalum Energetics. While the essentials of the niobium

system are understood, the energetics of the tantalum rearrange-

ments pose several problems: (1) since the olefin complexes

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4825

A R T I C L E S

Hirsekorn et al.

Figure 7. Relative standard free energies (kcal/mol, 103 °C) of (silox)₃Ta(ole) (2-ole) indicated by bottom of solid green column and those of (silox)₃-

Ta(alk) (2dalk) indicated by bottom of striped red column. The common bottom of the striped green and solid red columns indicates the relative ∆G° of

(silox)₂RTa(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-R). The solid red, red-striped, green, and green-striped columns represent the activation free energies associated

with k₁, k_-1, k₂, and k_-2of Scheme 7, respectively. The top of the two green columns is the relative standard free energy of the transition state for 2-ole to

6-R, and the top of the red columns is the ∆G° of the transition state for 6-R to 2dalk. All energies are relative to (silox)₃Ta(η-C₂H₄) (2-C₂H₄) at 0.0

kcal/mol, and the relative 2-ole energies are taken from computations. The arrows indicate that the respective state energies are maxima, and the bars below

the arrows indicate estimates of their probable values. See text for explanation.

Now in every case, save ethylene and norbornene, the second

toward the third-row metal. Likewise, the second-row niobium

should accommodate the pseudo-lower oxidation states of 1-ole

and 1dalk better than the tantalum congeners. The shift in

reaction coordinate relative to 2-ole and 2dalk that accompanies

the energy lowering of 6-R relative to its niobium counterpart

is difficult to comment about. It is plausible that tantalum

alkylidene species are more “M(V)-like” than niobium alkyl-

idene complexes, and that geometry changes favoring a more

compressed RC along the 6-R to 2dalk trajectory reflect that,

but absent substantial structural data this must remain specula-

tion. Because of the complicated nature of 6-R, no attempt to

model its energetics by quantum calculations was made.

Figure 6 can be used to generalize the tantalum system as

much as possible. It is expected that much of the energy differ-

ences in the ground states of (silox)₃Ta(ole) (2-ole) translate to

TS₁, because the metal olefin interaction is being disrupted in

this transition state. The activation free energies corresponding

transition state is lower than the first, despite ∆G₂^qbeing greater

q

than ∆G₁in every remaining case except cyclopentene.

It is the influence of ∆G°((silox)₂RTa(κ²-O,C-OSi^tBu₂CMe₂-

CH₂) (6-R)) and its reaction coordinate relative to (silox)₃-

Ta(ole) (2-ole) and (silox)₃Ta(alk) (2dalk) that are responsible

for the more rapid reaction rates in the tantalum system and

the corresponding transition state energy changes. Figure 6

provides a rough illustration of the energetics of the olefin to

alkylidene rearrangement, with the caveat that niobium and

tantalum have been placed on the same diagram only as a matter

of convenience. A natural consequence of the high energy

position of (silox)₂RNb(κ²-O,C-OSi^tBu₂CMe₂CH₂) (4-R) is that

variation of (silox)₃Nb(ole) (1-ole) ground-state energies results

in dramatic rearrangement rate changes because TS₂is rate-

determining. The significant stabilization of 6-R relative to its

niobium congener does not explain why TS₂is typically lower

in energy than TS₁in most tantalum cases. The change in

reaction coordinate for 6-R shown in Figure 6 must accompany

its energetic change relative to niobium to rationalize this

observation.

q

to k₁should not vary a great deal, and ∆G₁(ave) ) 29.3(16)

q

kcal/mol; with ∆G₁(2-CH₂CH₂Ph) removed, even less variation

is noted (28.7(6) kcal/mol). Likewise, the spread in ∆G^q

1

(without the outlying styrene case) is only 1.7 kcal/mol. There

q

is essentially no LFE relationship between ∆G₁and ∆G°(6-

From the standpoint of energies, the greater importance of

the tuck-in alkyl intermediate to the tantalum system may be

due to the true M(V) character of this intermediate. The olefin

and alkylidene species can be considered M(III)s understanding

the olefin complex certainly has metalacyclopropane character

and alkylidenes are usually treated as dianionicsand hence a

greater propensity toward the higher oxidation state should lie

q

R/2-ole); for ∆∆G₁) â∆∆G°(6-R/2-ole) + c₁, â < 0.2 with

or without inclusion of data pertaining to 2-CH₂CH₂Ph, and the

correlation coefficients are <0.1 in both cases.

Greater variation in activation free energies would be expected

for k_-1, since it is unlikely that variation in tuck-in alkyl energies

would be as strongly coupled to the olefin-like transition state

9

4826 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

A R T I C L E S

Figure 8. Relative standard free energies (kcal/mol, 155 °C) of (silox)₃Ta(η-C₂H₃C₆H₄-p-X) (2-C₂H₃C₆H₄-p-X, X ) CF₃(red), H (black), OMe (green)),

(silox)₂(X-p-C6H4CH2CH2)Ta(κ²-O,C-OSi^tBu₂CMe₂CH₂) (6-C₂H₄C₆H₄-p-X, X ) CF₃, H, OMe), and (silox)₃TadCHCH₂C₆H₄-p-X (2-C₂H₃C₆H₄-p-X, X

) CF₃, H, OMe). Activation energies for k₁, k_-1, k₂, and k₃of Scheme 7 are indicated by red (X ) CF₃), black (X ) H), and green (X ) OMe) dashed lines.

The tuck-in alkyls 6-C₂H₄C₆H₄-p-X have been arbitrarily placed at 0.0 kcal/mol, and all ground and transition state energies are given relative to this

reference state. See text for rationale.

TS₁. For the following arguments, ∆G°(6-R) values are

estimated on the basis of the assumption that 2-ole h 2dalk is

the same as that in the corresponding niobium cases as explained

above (the maximum ∆G_-1^qof 6-^cC₇H₁₁is estimated to be 27.8

kcal/mol; this corresponds to ∆G°(6-^cC₇H₁₁) being at least 3.4

kcal/mol greater than ∆G°(2d^cC₇H₁₀)). With the obviously

mol and ∆G_-2^q(ave, acyclic) ) 30.6(5) kcal/mol, with corre-

sponding ranges of 0.8 and 0.9 kcal/mol, respectively. The cyclic

q

cases display a ∆G₂(ave, cyclic) of 28.0(2) kcal/mol with a

small range (e0.3 kcal/mol), but possess a more scattered array

of G_-2^qvalues (29.6(14) kcal/mol) with a large range (2.6 kcal/

mol). This discrepancy from the view in Figure 6 is due to the

norbornene case, in which ∆G°(6-^cC₇H₁₁) is estimated.

q

outlying styrene data removed, ∆∆G_-1) â∆∆G°(2-ole/6-R)

+ c_-1and â < 0.94 with a correlation coefficient of R²) 0.97.

A correspondingly large ∆G_-1^q(ave) of 31.3(27) kcal/mol is

observed along with a range of ∆G_-1^q’s of 6.3 kcal/mol. It is

not surprising that 2-CH₂CH₂Ph and its TS₁are an oddity in

this assessment. Recall that reversible cyclometalation into the

R-position of the bound styrene was suspected from the labeling

studies, and it is postulated that the inductively withdrawing

phenyl substituent renders the olefin more strongly bound than

other acylic olefins. As espoused for the niobium-styrene case,

the reaction coordinate should be correspondingly elongated and

TS₁should be higher in energy than other acyclic cases.

As Scheme 7 shows, the transition from olefin complex to

tuck-in alkyl requires elongations of d(Ta-C_â) and d(silox

methyl C-H) as new Ta-C (from cyclometalation) and C_âH

bonds form. The reaction coordinate has a similar complexity

in the transition from tuck-in alkyl to alkylidene. The role the

reaction coordinate plays in distinguishing the energetics of the

various substrates of this study is difficult to understand without

structural information, but electronic factors determined in the

para-substituted styrene derivatives can provide some insight.

Figure 8 illustrates the energetics of (silox)₃Ta(η-C₂H₃Ph-p-X)

(X ) OMe, 2-C₂H₃Ph-p-OMe; H, 2-C₂H₃Ph; CF₃, 2-C₂H₃Ph-

p-CF₃) and their conversions to (silox)₂(X-p-PhCH₂CH₂)Ta(κ²-

O,C-OSi^tBu₂CMe₂CH₂) (X ) OMe, 6-C₂H₄Ph-p-OMe; H,

6-C₂H₄Ph; CF₃, 6-C₂H₄Ph-p-CF₃) and (silox)₃TadCHCH₂Ph-

p-X (X ) OMe, 2dCHCH₂Ph-p-OMe; H, 2dCHCH₂Ph; CF₃,

2dCHCH₂Ph-p-CF₃). The diagram is referenced to the three

6-R species at 0.0 kcal/mol at the same position along the

q

While LFE relationships appear to be present for ∆G₂and

∆G_-2^q, a close inspection revealed that the data are bunched in

acyclic and cyclic groupings and the fits are inappropriate.

Figure 6 suggests that for like species, the activation free

energies leading to the second transition state, TS₂, should be

q

similar. For the acyclic cases, ∆G₂(ave, acyclic) ) 32.8(4) kcal/

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4827

A R T I C L E S

Hirsekorn et al.

reaction coordinate. There is a modest check on this assumption

via some corresponding niobium equilibria. As Table 3 shows,

the ∆G₁₀₃°(1-C₂H₃Ph-p-OMe/1-C₂H₃Ph) value is 0.55 kcal/mol,

close to the 0.4 kcal/mol for the tantalum equilibrium based on

6-R as the reference state. In addition, the ∆G₁₀₃°(1-C₂H₃Ph-

p-CF₃/1-C₂H₃Ph) value is -1.22 kcal/mol, and Figure 8 shows

the related tantalum equilibrium to be -1.0 kcal/mol.

few direct measurements and calculated relative energies. For

the cyclics, 6-R is stable enough to be observable except for

6-^cC₇H₁₁, which is presumed to have steric interactions (Ta-

C(sp³)) that are more severe than 2d^cC₇H₁₀(Ta-C(sp²)).

Conclusions

Although the group 5 complexes above lack the space

necessary to do productive olefin metathesis chemistry, they

are representative of the class of compounds used as catalysts.

In these systems, metal alkylidenes that possess â-hydrogens

are not only kinetically stable, but also are most often more

thermodynamically stable than their olefin isomers. While this

contradicts some studies pertaining to later transition metals,

energetic differences between isomeric alkylidene and olefin

complexes are likely to be minimal. From a mechanistic

standpoint, the isomerization process required a substantial

amount of activation energy in every case, and each was studied

in nonpolar, aprotic media because of the sensitivity of the

compounds. Even though the siloxide ligand mediated the

reversible olefin to alkylidene rearrangements, a great deal of

reorganization energy was expended to form the cyclometalated

intermediate. It can be concluded on the basis of these studies

that â-hydrogen-substituted alkylidenes are thermodynamically

and kinetically quite stable. However, one can imagine that

alkylidene complexes that are stable to a variety of functional

groups, including those of polar and protic solvents, may be

subject to a number of isomerization pathways (e.g., L_nMd

CR(CH₂R′) + H⁺h [L_nM-CRH(CH₂R′)]⁺h L_nM(RHCd

CHR′) + H⁺) that do not have energy requirements of large

magnitude. In such instances, it may be the thermodynamic

stability of the alkylidenes that renders them catalytically active

under conditions where isomerization could severely inhibit

catalysis.

Finally, many olefin metathesis catalysts are generated in situ

via the combination of a high oxidation state complex, typically

a d⁰metal halide, and a compound that can serve as an alkylating

agent (e.g., MX_n+ 2/m R_mM′ f X_n-2MR₂+ 2/m X_mM′).^17,18

Following alkylation, R-abstraction can then lead to an alkyl-

idene (e.g., X_n-2MR₂f X_n-2MdCHR′ + RH) that is competent

for metathesis. An alternative pathway is suggested by these

investigations, whereby the high oxidation state transition metal

complex (MX_n) is reduced by the main group M′R_mspecies.

The lower valent compound may capture an olefin, and its

subsequent rearrangement to an alkylidene then engenders

metathesis.

Note that the biggest spread in energies corresponds to the

ground states of 2-C₂H₃Ph-p-X. If the reaction coordinates for

all X’s were the same, the TS₁’s for the X ) CF₃, H, and OMe

cases would be in that order, but they are reversed. The highest

energy olefin complex, 2-C₂H₃Ph-p-OMe, possesses the lowest

energy TS₁. The reaction coordinate for the conversion of

2-C₂H₃Ph-p-OMe to 6-C₂H₄Ph-p-OMe must be compressed

relative the X ) H, which in turn is compressed relative to X

) CF₃. While structural evidence is not available, one would

expect the d(TaC) of the more withdrawing CF₃-substituted

styrene to be shorter, followed by X ) H and finally X ) OMe.

This would correspond to a lengthening of the RC for X )

CF₃etc. A compression of the RC for X ) H, OMe is also

seen in the 6-C₂H₄Ph-p-X to 2dCHCH₂Ph-p-X conversion

through TS₂, although here the effect is less easily rationalized.

With ∆G°(6-C₂H₄Ph-p-X) serving as the reference state, the

fact that ∆G°(2dCHCH₂Ph-p-X) manifests little difference

makes sense because the para-substituents are not electronically

coupled to the metal. On the basis of Figure 6, minimal TS₂

energy differences are expected, and only an ∼0.8 kcal/mol

range is observed, but the substituents are reversed, and while

a compression in RC can explain this, its origin is not especially

transparent.

One oddity that deserves comment is the ∆G°(2-^cC₅H₈h

2d^cC₅H₈) of -1.3 kcal/mol in comparison to the -5.1 kcal/

mol observed for ∆G°(1-^cC₅H₈h 1d^cC₅H₈). Given the near

size equivalence of niobium and tantalum (r_cov) 1.34 Å), these

reproducible results are difficult to interpret. While not as large

a difference, the ∆G°(2-^cC₇H₁₀h 2d^cC₇H₁₀) of -1.7 kcal/

mol is also distinctly less than its niobium counterpart, ∆G°-

(1-^cC₇H₁₀h 1d^cC₇H₁₀), which has a value of -3.9 kcal/mol.

These disparate energies between second- and third-row species

hint at an unusual electronic origin. Since the metal alkylidene

bonds of the norbornene and cyclopentene species contain a

greater degree of s-character due to the modest ring strain in

these compounds, it is conceivable that subtleties in the overlap

2

and energies of the d_zorbitals of Nb and Ta may manifest

2

themselves here. It has been shown calculationally that the d_z

Experimental Section

orbital of (HO)₃Ta (2′) is substantially lower than the corre-

sponding niobium orbital,³⁸and this is, in part, what is

responsible for the dramatic stability and reactivity of (silox)₃Ta

(2)³⁹relative to its nonisolable counterpart, (silox)₃Nb (1).

General Considerations. For a detailed experimental, see the

Supporting Information. All manipulations were performed using either

glovebox or high vacuum line techniques. All glassware was oven-

dried, NMR tubes for sealed tube experiments were flame-dried under

vacuum, all solvents were scrupulously dried and degassed, and all

gases were dried. (silox)₃Ta (1-Ta),⁴⁰(silox)₂HTa(κ²-O,C-OSi^tBu₂CMe₂-

CH₂) (8),⁴²(silox)₃Ta(η-H₂CCHR) (R ) H, 2-C₂H₄; Me, 2-C₂H₃Me),⁴²

(silox)₃Nb(4-pic) (1-4-pic),^37,38(silox)₃Nb(η-H₂CCHPh) (1-C₂H₃Ph),³⁷

(silox)₃NbPMe₃(1-PMe₃),³⁸â,â-dideuteriostyrene,⁷⁷2,3-dideuterio-

norbornene,^78-80and 1,2-dideuteriocyclohexene⁸³were prepared by

literature procedures.

In summary, as in the niobium system, the ground-state

energies of the olefin complexes, (silox)₃Ta(ole) (2-ole), play

a major role in affecting the energies of TS₁. This step is isolated

for all substrates except norbornene, and very modest changes

in rate are found for the formation of the tuck-in alkyls,

(silox)₂RTa(OSi^tBu₂CMe₂CH₂) (6-R), which is consistent with

the view in Figure 6. The energy and position along the reaction

coordinate of 6-R change the nature of the rearrangement for

tantalum. Relative to 2-ole and 2dalk, 6-R is more stable in

the acyclic cases,⁷⁵although 2-ole h 2dalk is probably very

similar to that of the niobium system, at least according to a

Procedures. 1. (silox)₃NbCl₂(3). Into a 100-mL glass bomb charged

with NbCl₅(4.49 g, 16.62 mmol) and Na(silox)⁹⁷(12.10 g, 50.75 mmol,

(97) Covert, K. J.; Wolczanski, P. T.; Hill, S. A.; Krusic, P. J. Inorg. Chem.

1992, 31, 66-78.

9

4828 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

(silox)₃M(ole) to (silox)₃M(alk) Rearrangements

A R T I C L E S

3.05 equiv) 20 mL of toluene was distilled at -78 °C. The bomb was

sealed and heated to 100 °C for 10 d, giving a pale yellow-green slurry.

Upon filtration, crystallization from THF at -78 °C yielded 11.85 g

of colorless 3 in two crops (75%). Anal. Calcd for C₃₆H₈₁Si₃O₃NbCl₂:

C, 53.38; H, 10.08. Found: C, 53.29; H, 10.30.

(η-^cC₇H₁₀) (2-^cC₇H₁₀). After being stirred for 12 h at 23 °C, 145 mg

of orange 2-^cC₇H₁₀was crystallized from diethyl ether (47%). Anal.

Calcd for C₄₃H₉₁Si₃O₃Ta: C, 56.03; H, 9.95. Found: C, 55.90; H, 9.85.

e. (silox)₃Ta(η-C₂H₃Ph) (2-C₂H₃Ph). After being stirred for 1 h at 23

°C, 195 mg of orange 2-C₂H₃Ph was crystallized from pentane (65%).

Anal. Calcd for C₄₃H₉₁Si₃O₃Ta: C, 56.74; H, 9.63. Found: C, 56.05;

H, 9.64. f. (silox)₃Ta(η-C₂H₃C₆H₄-p-OMe) (2-C₂H₃Ph-p-OMe). After

being stirred for 1 h at 23 °C, 195 mg of red 2-C₂H₃Ph-p-OMe was

crystallized from pentane (62%). g. (silox)₃Ta(η-C₂H₃C₆H₄-p-CF₃) (2-

C₂H₃Ph-p-CF₃). After being stirredfor 1 h at 23 °C, 180 mg of orange

2-C₂H₃Ph-p-CF₃was crystallized from pentane (55%).

6. (silox)₃NbdCHⁿPr (1dCHⁿPr). A 50-mL bomb charged with

350 mg of 1-C₂H₃Et (0.440 mmol) and 20 mL of benzene was heated

at 155 °C for 8.5 h. Upon cooling a diethyl ether solution to -78 °C,

90 mg of microcrystalline red 1dCHⁿPr was obtained (26%). Anal.

Calcd for C₄₀H₈₉Si₃O₃Nb: C, 60.41; H, 11.28. Found: C, 60.19; H,

11.42.

2. (silox)₃Nb(ole) (1-ole) from (silox)₃Nb(4-pic) (1-4-pic). A 50-

mL bomb reactor was charged with (silox)₃Nb(4-pic) (1-4-pic), the

system was evaculated, and 10 mL of benzene was transferred. The

gaseous reagent (∼5 equiv) was condensed into the bomb at 77 K from

a calibrated gas bulb. The reaction mixture was stirred for variable

times and temperatures, transferred to a flask, and crystallized. a.

(silox)₃Nb(η-C₂H₄) (1-C₂H₄). After a 30-min reaction time at 23 °C,

isolation from diethyl ether yielded 333 mg of 1-C₂H₄as dark green

microcrystals (92%). b. (silox)₃Nb(η-C₂H₃Me) (1-C₂H₃Me). After a

30-min reaction time at 23 °C, isolation from diethyl ether yielded 250

mg of 1-C₂H₃Me as dark green microcrystals (67%). Anal. Calcd for

C₃₉H₈₇Si₃O₃Nb: C, 59.96; H, 11.22. Found: C, 59.20; H, 10.89. c.

(silox)₃Nb(η-C₂H₃Et) (1-C₂H₃Et). After a 30-min reaction time at 23

°C, isolation from diethyl ether yielded 385 mg of 1-C₂H₃Et as dark

green microcrystals (81%). Anal. Calcd for C₄₀H₈₉Si₃O₃Nb: C, 60.41;

H, 11.28. Found: C, 60.25; H, 11.34. d. (silox)₃Nb(η-cis-C₂H₂Me₂)

(1-cis-C₂H₂Me₂). After heating for 16 h at 85 °C, isolation from pentane

afforded 75 mg of dark green 1-cis-C₂H₂Me₂(40%). e. (silox)₃Nb(η-

C₂H₃C₆H₄-p-OMe) (1-C₂H₃Ph-p-OMe). After being stirred for 18 h

at 23 °C, isolation from diethyl ether afforded 170 mg of green-brown

1-C₂H₃Ph-p-OMe (46%). Anal. Calcd for C₄₅H₉₁Si₃O₄Nb: C, 61.89;

H, 10.50. Found: C, 61.70; H, 10.71. f. (silox)₃Nb(η-C₂H₃C₆H₄-p-

CF₃) (1-C₂H₃Ph-p-CF₃). After being stirred for 18 h at 23 °C, isolation

from diethyl ether afforded 170 mg of army green 1-C₂H₃Ph-p-CF₃

(44%).

7. (silox)₃Nbd^cC₆H₁₀(1d^cC₆H₁₀). A 50-mL bomb charged with 600

mg of 1-^cC₆H₁₀(0.732 mmol) and 25 mL of benzene was heated at 55

°C for 13 d. Upon cooling a pentane solution to -78 °C, 415 mg of

dark green microcrystalline 1d^cC₆H₁₀was obtained (69%). Anal. Calcd

for C₄₂H₉₁Si₃O₃Nb: C, 61.60; H, 11.90. Found: C, 60.91; H, 11.84.

n

8. (silox)₃BuTa(K²-O,C-OSi^tBu₂CMe₂CH₂) (6-ⁿBu). An NMR tube

was charged with 180 mg of 1-C₂H₃Et (0.204 mmol) and 1 mL of

toluene-d₈. After being heated for 12 h at 103 °C, the solution was

pale yellow. The solvent was allowed to evaporate over the course of

7 d to give 90 mg of 6-ⁿBu (50%) as a crystalline solid. Anal. Calcd

for C₄₀H₈₉Si₃O₃Ta: C, 54.39; H, 10.16. Found: C, 53.86; H, 9.69.

NMR Tube Reactions. Flame-dried NMR tubes, sealed to 14/20

ground glass joints, were charged with the organometallic reagent

(typically 20 mg) and any other solid reagent in a drybox and removed

to the vacuum line on needle valve adapters. The NMR tube was

degassed, and after transfer of deuterated solvent, a calibrated gas bulb

was used to introduce volatile reagents at 77 K if necessary. The tubes

were then sealed with a torch. Sometimes freshly distilled solvent was

added to the tubes while in the drybox, then moved to the vacuum

line, degassed, and sealed with a torch.

3. (silox)₃Nb(ole) (1-ole) from (silox)₃Nb(PMe₃) (1-PMe₃). Into a

flask charged with (silox)₃NbPMe₃(1-PMe₃), 25 mL of pentane and

excess olefin were distilled. The reaction mixture was stirred for variable

times and temperatures, transferred to a flask, and crystallized. a.

t

(silox)₃Nb(η-C₂H₃Bu) (1-C₂H₃Bu). After stirring 1-PMe₃with ∼20

t

equiv of C₂H₃Bu for 2 h at 23 °C, isolation from diethyl ether afforded

t

342 mg of brown microcrystalline 1-C₂H₃Bu (90%). Anal. Calcd for

C₄₂H₉₃Si₃O₃Nb: C, 61.27; H, 11.39. Found: C, 60.23; H, 11.33. b.

(silox)₃Nb(η-^cC₅H₈) (1-^cC₅H₈). After stirring 1-PMe₃with ∼35 equiv

of ^cC₅H₈for 2 h at 23 °C, isolation from diethyl ether afforded 130 mg

of green microcrystalline 1-^cC₅H₈(50%). Anal. Calcd for C₄₁H₈₉Si₃O₃-

Nb: C, 60.00; H, 11.11. Found: C, 61.06; H, 11.13.

Kinetics Studies. Sets of three and six NMR tubes for kinetics

monitoring of the olefin to alkylidene isomerization were prepared as

described above. Once the solution was added to the NMR tubes, they

were transferred to the vacuum line attached to a 3- or 6-prong adapter,

evacuated and sealed. Kinetics experiments were typically done in sets

of three for adequate statistics. The sets of NMR tubes were placed

simultaneously in a temperature-controlled oil bath for a measured

amount of time and periodically removed, frozen at 77 K, and warmed

4. (silox)₃Nb(ole) (1-ole) from (silox)₃NbCl₂(3). Into a flask charged

with 3 and 2.1 equiv of 0.65% Na/Hg, THF and excess olefin were

condensed at 77 K. After being stirred for 12 h, the THF was removed

and replaced with hydrocarbon solvent. The solution was filtered, and

the complex was crystallized. a. (silox)₃Nb(η-^cC₆H₁₀) (1-^cC₆H₁₀).

Crystallization from hexanes provided 600 mg of dark green 1-^cC₆H₁₀

(40%). Anal. Calcd for C₄₂H₉₁Si₃O₃Nb: C, 61.42; H, 11.17. Found:

C, 61.19; H, 11.02. b. (silox)₃Nb(η-^cC₇H₁₀) (1-^cC₇H₁₀). Crystallization

from diethyl ether afforded 253 mg of green microcrystalline 1-^cC₇H₁₀

(25%).

1

and analyzed by H NMR spectroscopy. For the specific resonances

monitored, see Supporting Information.

Kinetics Modeling. All experimental data were analyzed using

Scientist 2.0 by MicroMath for Microsoft Windows. Least squares

regression analyses were employed to model all forward and reverse

rate constants as well as activation parameters via the Eyring equation.

Equilibrium Studies. General. NMR tubes for measuring the

equilibria listed in Table 3 were prepared as described above. After

5. (silox)₃Ta(ol) (2-ole). General. A flask or bomb reactor (in the

case of gaseous olefins) was charged with (silox)₃Ta (2), a hydrocarbon

solvent, and an excess of olefin. After being stirred for a period of

time at 23 °C, the blue color changed to orange, the solvent was

removed, and the yellow-orange to red solid was triturated, dissolved

in hexanes, and filtered. Crystallization was typically from hexanes or

diethyl ether at -78 °C. a. (silox)₃Ta(η-C₂H₃Et) (1-C₂H₃Et). After

being stirred for 1 h at 23 °C, 200 mg of orange 1-C₂H₃Et was

crystallized from pentane (67%). Anal. Calcd for C₄₀H₈₉Si₃O₃Ta: C,

54.39; H, 10.16. Found: C, 54.23; H, 10.17. b. (silox)₃Ta(η-^cC₅H₈)

(2-²C₅H₈). After being stirred for 1 h at 23 °C, 222 mg of orange

2-^cC₅H₈was crystallized from diethyl ether (63%). c. (silox)₃Ta(η-

^cC₆H₁₀) (2-^cC₆H₁₀). After being stirred for 24 h at 23 °C, 186 mg of

orange 2-^cC₅H₈was crystallized from diethyl ether (55%). d. (silox)₃Ta-

1

equilibrium was achieved, integration of appropriate H NMR spec-

troscopic resonances afforded equilibrium constants. Specific resonances

analyzed are given in the Supporting Information.

Single-Crystal X-ray Diffraction Studies. A crystal was isolated,

covered in polyisobutylene, and placed under a 173 °C N₂stream on

the goniometer head of a Siemens P4 SMART CCD area detector

system (graphite-monochromated Mo KR radiation, λ ) 0.71073 Å).

The structure was solved by direct methods (SHELXS). All non-

hydrogen atoms were anisotropically refined, and hydrogen atoms were

treated as idealized contributions.

9. (silox)₃Nbd^cC₆H₁₀) (1-^cC₆H₁₀). Dark green crystals of 1-^cC₆H₁₀

were grown from evaporating a concentrated benzene-d₆solution. In

space group Pca2₁, correlation problems are known to interfere with

9

J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005 4829

A R T I C L E S

Hirsekorn et al.

normal refinement if a local center of symmetry is near x ) 1/8 and y

) 1/4.⁹⁸Two local centers of symmetry at x ) 0.85, y ) 0.749 and x

) 0.860, y ) 1.249 were found. Even greater problems may arise if

some of the atoms have special values for y (0.0 and 0.5 for a pair of

related atoms).⁹⁸Note that the atom coordinates of the niobiums meet

these criteria (atom, x, y, z: Nb1, 0.7468, 0.50578, 0.345; Nb1A, 0.9614,

0.9921, 0.4100; Nb1B, 0.7607, 1.0057, 0.6587; Nb1C, 0.9614, 1.4929,

0.5973). In this structure, eight of nine tert-butyl groups are related by

a local inversion center; hence, while the structure is correctly refined

in a noncentrosymmetric space group, it is very nearly centrosymmetric,

leading to some problems of correlation. Refinement in the centrosym-

metric space group Pbca afforded an R₁of ∼16%. The average values

of bond lengths and angles are reported for this molecule because they

should be unaffected by this pseudo-symmetry.

for hydrogen. This level of theory was selected on the basis of a series

of test calculations on the singlet and triplet states of Nb(OH)₃(1′),

Ta(OH)₃(2′),³⁸and their olefin adducts.

Acknowledgment. This research is dedicated to the memory

of Ian P. Rothwell, a great friend and colleague, and a master

of group 5. Support from the National Science Foundation

(P.T.W., CHE-0212147; T.R.C., CHE-0309811) and Cornell

University is gratefully acknowledged. Profs. Barry K. Carpen-

ter, Charles F. Wilcox, David B. Collum, Kit Cummins, and

Richard Marsh are thanked for helpful discussions.

Supporting Information Available: Crystallographic data

(CIF), a detailed experimental description, and a discussion of

entropy of activation estimations (PDF). This material is

available free of charge via the Internet at http://pubs.acs.org.

10. (silox)₃Ta(η-C₂H₃Et) (2-C₂H₃Et). Bright orange crystals of

2-C₂H₃Et were grown from evaporating a concentrated benzene solution.

11. (silox)₃BuTa(K²-O,C-OSi^tBu₂CMe₂CH₂) (6-ⁿBu). A light

n

JA046180K

yellow crystal of 6-ⁿBu from procedure 8 was used. The tert-butyl

groups of one silox (Si3) were disordered, and each was treated as two

units of half-occupancy.

(99) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.

A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann,

R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,

K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,

R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;

Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;

Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz,

J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;

Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,

C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;

Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,

M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision x.x; Gaussian,

Inc.: Pittsburgh, PA, 1998.

Calculations. To obtain the minima in this research, full geometry

optimizationsswithout any metric or symmetry restrictionsswere

performed using the Gaussian⁹⁹package, and these employed density

functional theory, specifically the BLYP functional.¹⁰⁰Atoms were

described with the Stevens effective core potentials and attendant

valence basis sets.¹⁰¹This scheme, dubbed CEP-31G(d), entails a

valence triplet ú description for the transition metals, a double-ú-plus-

polarization basis set for main group elements, and the -31G basis set

(100) Parr, R. G.; Yang, W. Density-Functional Theory of Atoms and Molecules;

Oxford University Press: Oxford, 1989.

(98) Marsh, R. R.; Schomaker, V.; Herbstein, F. H. Acta Crystallogr. 1998,

B54, 921-924.

(101) Stevens, W. J.; Krauss, M.; Basch, H.; Jasien, P. G. Can. J. Chem. 1992,

70, 612-630.

9

4830 J. AM. CHEM. SOC. VOL. 127, NO. 13, 2005

Article Doi

Thermodynamics, kinetics, and mechanism of (silox)3M(olefin) to (silox)3M(alkylidene) rearrangements (silox = tBu 3SiO; M = Nb, Ta)

DOI: 10.1021/ja046180k

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja046180k

Full text of DOI:10.1021/ja046180k

Products guided by the article

R&D Labs maybe for 851267-60-2

Relevant to this article

Hot Product